final architecture for tvws spectrum sharing systems - cogeu

COGEU

FP7 ICT-2009.1.1

COgnitive radio systems for efficient sharing of TV white spaces

in EUropean context

COGEU D3.3

Final Architecture for TVWS Spectrum Sharing Systems

Contractual Date of Delivery to the CEC: September 2011

Actual Date of Delivery to the CEC: November 2011

Author(s): Damien Lavaux (TC), Paulo Marques (IT), Joseph Mwangoka (IT), Jorge Ribeiro

(IT), Álvaro Gomes (PTI), Helder Alves (PTI), Carlos Silva (PTI), Evagoras Charalambous

(SIGINT)

Participant(s): IT, TC, PTI, SIGINT

Workpackage: WP3

Est. person months: 21

Security: Public

Nature: Report

Version: v1.0

Total number of pages: 97

Abstract:

This Deliverable is related to tasks T3.3 “COGEU Reference Architecture for Spectrum Commons” and T3.4 “COGEU Reference Architecture for Secondary Spectrum Market”. It provides an overview of system architectures proposed by TV White-space standards, present the COGEU vision on spectrum sharing and presents the COGEU Reference Architecture that is used as a basis for cross-work-packages discussions.

Keyword list: TVWS, Architecture, Design, Dynamic Spectrum Management, Secondary spectrum trading, spectrum broker, regulatory scenarios, COGEU reference model, TVWS allocation, Geolocation Database.

COGEU D3.3 – Architecture for TVWS Spectrum Sharing Systems

of 97

Executive Summary

As stated since the beginning of the project, COGEU is dedicated to investigate both Spectrum Commons and Secondary Spectrum Trading approaches for the exploitation of the TV White Space. In the spectrum commons usage model, there is no spectrum management entity to preside over the resource allocation. Only permissions to access the spectrum is provided. The spectrum commons regime promotes sharing, but does not provide adequate Quality of Service in a world where applications are more and more resources demanding. For applications that require sporadic access to spectrum and for which QoS guarantees are important, temporary licensed spectrum with near automatic secondary markets is expected to be a relevant solution. Trading allows players to directly trade spectrum usage rights, thereby establishing a secondary market for spectrum leasing and spectrum auction. Of course, both regimes, spectrum commons and spectrum trading, are only possible to the extent allowed by national regulation. For that matter, COGEU considered two regulatory assumptions that have been carefully validated with both regulation bodies and driving technology providers:

1

st regulatory assumption: Both geo-location database access and spectrum sensing are required for

the protection of Incumbents. 2

nd regulatory assumption: Only geo-location database access is required for the protection of the

Incumbents. Before giving the details of the two scenarios correlated to those regulatory assumptions, and the respective reference architectures proposed, this document provides an extensive analysis of the current State-of-the-Art. By analysing current European regulation status, standardisation proposals and research literature/projects, COGEU consortium was led to the conclusion that: currently, there is no architecture proposed for secondary spectrum trading in the TV White-Spaces. All of them assume spectrum commons approach for TVWS access. A clear need has been identified to adapt the state of the art on architecture proposals in order to fully satisfy the COGEU requirements. More specifically, the topic of geo-location database driven access to TVWS has been adapted and extended by the project. The main innovation brought by COGEU is in the combination of unlicensed access to TV white spaces with secondary spectrum trading mechanisms based on the geo-location database (as shown in Figure 1), thanks to the 2

nd regulatory scenario. This tool enables regulators to dynamically

partition spectrum to be allocated to one or the other regime depending on socio-economics aspects.

Figure 1: Combination of unlicensed access to TVWS with secondary spectrum trading For both spectrum regimes the COGEU consortium proposed respective reference architecture in order to take into account both COGEU system requirements and the constraints introduced by the protection of the primary users. Under the Second regulatory scenario, which is the one that inspired the COGEU original proposal, a centralized topology with a spectrum broker trading with players is considered. The spectrum broker


of 97

controls the amount of bandwidth and power assigned to each user in order to keep the desired QoS and interference below the regulatory limits. In the COGEU reference model, the centralised Broker is an intermediary between a geo-location database (spectrum information supplier) and players that negotiate spectrum on behalf of spectrum users. Being the central entity in the proposed architecture, the COGEU spectrum broker have been fully specified in order to ensure a smooth implementation during the remaining work tasks. Features introduced have been derived into correlated internal modules of the COGEU broker. The main modules being:

TVWS Occupancy Repository: dealing with secondary users’ co-existence, and Primary user protection.

Spectrum Policies Repository: securing the policy management and distribution mechanisms, it also includes policy management tools

Trading Information Repository: responsible for past transactions retrieval Interfaces specification is also provided in order for the Broker to communicate with:

Secondary Users (seen as players)

Geolocation database: there is a two-way communication pattern in order to retrieve information or to be notified when TVWS availability change in the DB.

Regulators: in order for regulatory bodies to keep control of spectrum allocation and advertise specific regulatory policies.

A COGEU geo-location database is also defined based on section 1 overview analysis. It has to deal with two operation models. Indeed, the COGEU geo-location database receives enquires from both unlicensed TVWSD’s and the broker. Finally, based on the reference model and architecture components, three scenarios identified in D3.1 are instantiated. This exercise is intended to identify where does the specific needs lie for each application scenario and to map possible scenarios identified in COGEU to the derived Reference Architecture. LTE-over-TVWS scenario instantiation highlights possible steps and specific components interactions required in order benefits from TVWS access. The same work as been done for WiFi-over-TVWS scenario that can be implemented under the two regulatory scenario assumed in COGEU, while Public Safety over TVWS focuses on the specific needs of disaster relief and emergency situations requirements.


of 97

Table of Contents

1- INTRODUCTION ............................................................................................................................................ 7

2- OVERVIEW OF SYSTEM ARCHITECTURES PROPOSED BY TV WHITE-SPACE STANDARDS

AND OTHER EUROPEAN PROJECTS .............................................................................................................. 8

2.1- ECMA 392 ................................................................................................................................................. 8 2.2- IEEE P1900.4A ........................................................................................................................................... 9 2.3- IEEE 802.19 ............................................................................................................................................... 9 2.4- IEEE 802.22 ............................................................................................................................................. 10 2.5- IEEE 802.11AF (WHITE-FI) ...................................................................................................................... 11 2.6- ETSI RRS ................................................................................................................................................. 12 2.7- CEPT/ECC SE43 ...................................................................................................................................... 13 2.8- WEIGHTLESS PROTOCOL ........................................................................................................................... 13 2.9- IETF PAWS WORKING GROUP ................................................................................................................ 14 2.10- ICT-QOSMOS ..................................................................................................................................... 14 2.11- ICT-SACRA ......................................................................................................................................... 16 2.12- CROSS ANALYSIS FOR COGEU SYSTEM ARCHITECTURE ....................................................................... 17

3- COGEU APPROACH AND REGULATORY SCENARIOS .................................................................... 21

3.1- OVERVIEW OF COGEU SYSTEM REQUIREMENTS ...................................................................................... 21 3.2- REGULATORY SCENARIO 1: SPECTRUM COMMONS WITH SENSING AND GEO-LOCATION ACCESS REQUIRED

23 3.3- REGULATORY SCENARIO 2: COMMONS AND SECONDARY TRADING, ONLY GEO-LOCATION ACCESS

REQUIRED ............................................................................................................................................................. 24

4- ARCHITECTURE FOR SPECTRUM COMMONS WITH SENSING AND GEO-LOCATION

ACCESS REQUIRED ........................................................................................................................................... 26

4.1- MASTER-SLAVE ARCHITECTURE ............................................................................................................... 28 4.1.1- Master devices .................................................................................................................................. 28 4.1.2- Slave devices .................................................................................................................................... 29 4.1.3- Information exchange between geo-location database and WSD .................................................... 29

4.2- COOPERATIVE SENSING ARCHITECTURES ................................................................................................. 30 4.2.1- Protection of incumbent users .......................................................................................................... 30 4.2.2- Cooperative Sensing in Cognitive Radios ........................................................................................ 31 4.2.3- Cluster Based sensing Architecture ................................................................................................. 32 4.2.4- IEEE 802.22 Sensing overview ........................................................................................................ 32 4.2.5- Architecture for combination of geo-location database with sensing .............................................. 32 4.2.6- Methodology for combination sensing with geo-location database ................................................. 33 4.2.7- General Architecture Description .................................................................................................... 34

5- ARCHITECTURE FOR COMMONS AND SECONDARY TRADING, ONLY GEO-LOCATION

ACCESS REQUIRED ........................................................................................................................................... 37

5.1- TVWS ALLOCATION AND TRADING MECHANISM ...................................................................................... 40 5.1.1- TVWS Allocation .............................................................................................................................. 40 5.1.2- Trading mechanism .......................................................................................................................... 41

5.2- NEGOTIATION PROTOCOLS FOR TRADING SPECTRUM RIGHTS WITH THE BROKER ...................................... 43 5.2.1- Pricing mode protocol ...................................................................................................................... 43 5.2.2- Auction mode protocol ..................................................................................................................... 44

5.3- PAYMENT SYSTEM .................................................................................................................................... 45 5.4- EXTERNAL GEO-LOCATION SPECTRUM DATABASE .................................................................................... 46 5.5- COGEU BROKER INTERNAL REPOSITORIES ............................................................................................... 48

5.5.1- TVWS occupancy repository ............................................................................................................ 48 5.5.2- Trading Information Repository ....................................................................................................... 50 5.5.3- Policies Repository ........................................................................................................................... 51

5.6- GRAPHICAL USER INTERFACE ................................................................................................................... 57 5.7- INTERFACES FOR GEO-LOCATION DATABASE ACCESS AND REGULATORY ENFORCEMENT ......................... 58

5.7.1- Interfaces and Protocols for Geo-location database access ............................................................ 58 5.7.2- Interface for populating the geo-location database ......................................................................... 59

5.8- INTERFACE BETWEEN THE COGEU BROKER AND PLAYERS ...................................................................... 61


of 97

5.9- REGISTRATION AND VALIDATION IN THE COGEU BROKER ...................................................................... 63 5.10- MOBILITY MANAGEMENT ARCHITECTURE FEATURES TO SUPPORT SEAMLESS HANDOVERS IN

HETEROGENEOUS ENVIRONMENTS ........................................................................................................................ 66 5.10.1- Introduction ...................................................................................................................................... 66 5.10.2- System Requirements ........................................................................................................................ 68 5.10.3- Mobility Management Architecture Features................................................................................... 69 5.10.4- Overview of the MIH Framework .................................................................................................... 70 5.10.5- Handover Mechanism ...................................................................................................................... 73 5.10.6- Scenario 1 for Commons: Sensing and geo-location database required ......................................... 73 5.10.7- Commons Scenario 2: Only geo-location required .......................................................................... 76 5.10.8- Analysis of the impact of COGEU mobility management approaches to the system requirements . 78 5.10.9- Conclusion and future directions ..................................................................................................... 80

6- INSTANTIATION OF THE COGEU REFERENCE ARCHITECTURE ............................................... 81

6.1- LTE OVER TVWS ..................................................................................................................................... 81 6.1.1- COGEU Instantiation of LTE over TVWS ........................................................................................ 81 6.1.2- FDD and TDD.................................................................................................................................. 83

6.2- WIFI OVER TVWS ................................................................................................................................... 85 6.3- PUBLIC SAFETY OVER TVWS.................................................................................................................... 87

7- CONCLUSIONS AND FUTURE WORK ................................................................................................... 90

8- REFERENCES ............................................................................................................................................... 91


of 97

1- Introduction

This Deliverable is related to tasks T3.3 “COGEU Reference Architecture for Spectrum Commons “and T3.4 “COGEU Reference Architecture for Secondary Spectrum Market”. The main objectives of those tasks are to derive the COGEU reference architecture, and define the basic functional blocks and their interoperations to implement a real time spectrum management platform for both Spectrum Commons and Secondary Spectrum Market. This deliverable is arranged as follows. Chapter 2- presents an overview of current standards and an analysis of the architectural trends that are of particular interest for the COGEU architecture. The objective of this study is to align the proposed COGEU architecture with the ongoing worldwide standardization effort. Combined with the previous use-cases analysis available in D3.1, it is expected to identify technology adaptation needs corresponding to the specificity of COGEU. Chapter 3- begins with a reminder of the system requirements introduced by previous WP2 reports. It clearly highlights the link between prioritized features considered, and the current implementation in WP7. This chapter then introduces COGEU approach and regulatory scenarios. Since no reference model can be generic enough to fit to all possible European Regulatory rules, two main regulatory scenarios are defined and motivated to be used as a basis for discussions. Then Chapter 4- and Chapter 5- summarize the design proposals for COGEU reference architectures corresponding to the previously introduced regulatory scenarios. They respectively define functional blocks, interfaces, protocols and maps the research studies done in complementary work-packages to modules of the COGEU Reference Architecture. Finally, Chapter 6- can be understood as an exercise to validate COGEU design, by instantiating three application scenarios based on the Initial Reference Architecture. It does not mean to be a complete mapping of functionality derived from the initial architecture, but it is intended to identify where does the specific needs lies for each application scenario.


of 97

2- Overview of system architectures proposed by TV White-space standards and other European projects

This section provides an overview of current standards and an analysis of the architectural trends that are of particular interest for the COGEU architecture. The objective of this study is to align the proposed COGEU architecture with the ongoing worldwide standardization effort. We introduce the activities in Cognitive Radio research communities and industry related to architecture and spectrum management, such as IEEE 802.22, ECMA 392, and IEEE 802.11af (also known as Wi-Fi 2.0 or White-Fi).

2.1- ECMA 392

This Standard specifies a medium access control (MAC) sub-layer and a physical (PHY) layer for personal/portable cognitive wireless networks operating in TV bands. This Standard also specifies a MUX sub-layer for higher layer protocols. It also specifies a number of incumbent protection mechanisms which may be used to meet regulatory requirements (which are out of the scope of ECMA392). ECMA 392 is the first Cognitive Radio standard for personal/portable devices to exploit the TVWS. It was started by the Cognitive Networking Alliance (CogNeA), and a draft specification was later transferred to TC48-TG1. The standard specifies PHY and MAC layers with several characteristics: flexible network formation, adaptation to different regulatory requirements, and support for real-time multimedia traffic. Ecma 392 is expected to enable new applications using TVWS such as in-home HD video transmission, campus-wide wireless coverage, and interactive TV broadcasting services. Ecma 392 has potential to deliver high-quality WS services in developed areas. However, such populated regions will introduce a more challenging environment for PU protection due to the high volume of DTV receivers and WMs. Therefore, it is important to determine safe operational conditions according to various service scenarios. TVWS may also create an interference-prone environment between neighbouring CRNs due to the characteristics of TV bands offering wider coverage, unlike the 60 GHz band in-home networks targeting short-range communications

Figure 2 gives an example scenario for the Ecma 392 standard. A master device as defined in this standard will meet the requirements of the FCC defined Mode II device by including geo-location (and sensing) functions and periodically obtaining available channels list from an authorized spectrum database via the internet. All slave devices (with sensing function) associated with such a master device will comply with the requirements of an FCC defined Mode I device.

Figure 2 : White Spaces Network


of 97

Relevance to COGEU Architecture Scenario While the definition of a MAC and PHY layer is out of the scope of this WP, assumptions on the nature of those layers are important in order to define a relevant architecture. Moreover, scenario considered in ECMA 392 paved the way to the definition of more complex scenario introduced in COGEU.

2.2- IEEE P1900.4a

IEEE P1900.4 is a working group within SCC41 (Standards Coordinating Committee 41) aiming at defining “Architectural Building Blocks Enabling Network-Device Distributed Decision Making for Optimized Radio Resource Usage in Heterogeneous Wireless Access Networks”. A standard, IEEE 1900.4-2009, has been published in February 2009. This standard defines the architectural building blocks, the interfaces, the information model and the procedures for optimized radio resource usage in heterogeneous wireless access networks. Three use cases are addressed by the IEEE 1900.4-2009 architecture:

Dynamic Spectrum Assignment: frequencies are dynamically assigned to Radio Access Networks (RAN);

Dynamic Spectrum Sharing: frequency bands assigned to RANs are fixed but a given band is potentially shared between several RANs;

Dynamic Radio Resource Usage Optimization: terminals choose, in a distributed manner, which radio access technology/technologies (RATs) to connect to.

On April 2009, P1900.4 working group has been working on two subsequent projects, namely P1900.4a and P1900.4.1. P1900.4a, entitled “Standard for Architectural Building Blocks Enabling Network-Device Distributed Decision Making for Optimized Radio Resource Usage in Heterogeneous Wireless Access Networks - Amendment: Architecture and Interfaces for Dynamic Spectrum Access Networks in White Space Frequency Bands”, aims at developing an amendment to the standard published in February 2009. The draft amendment consists in modifications the architecture, interfaces, information model and procedures of IEEE 1900.4-2009 for the P1900.4a system to operate in white space frequency bands. The P1900.4a system aims at enabling the coexistence of secondary systems operating in white spaces by providing the Base Stations (Cognitive Base Stations – CBSs – in the P1900.4a terminology) and the terminals with the following capabilities: spectrum sensing control, silent period management, white space classification and access to white space database. Additionally, the integration of a system operating in white space frequency bands into the heterogeneous wireless system of an operator is part of the P1900.4a use cases. In this use case, CBSs are capable of operating in white spaces and in licensed spectrum. The second project, P1900.4.1 aims at defining the protocols and the Service Access Points (SAPs) associated with the interfaces standardized in IEEE 1900.4-2009. Some of these protocols and SAPs are likely to be reused in the P1900.4a system. Relevance to COGEU architecture specification: The architecture developed in P1900.4a has a scope similar to the COGEU architecture since it addresses the coexistence of secondary systems and spectrum management in TV white space bands. Protocols defined in P1900.4.1 can be of interest to COGEU since they will be used in the system operating in white space frequency bands (developed in P1900.4a project).

2.3- IEEE 802.19

IEEE 802.19 is the Wireless Coexistence Technical Advisory Group (TAG) within the IEEE 802 LAN/MAN Standards Committee. The purpose of the standard is to enable the family of IEEE 802 Wireless Standards to most effectively use TV White Space by providing standard coexistence methods among dissimilar or independently operated TVWS networks and dissimilar TVWS Devices. This standard addresses coexistence for IEEE 802 networks and devices and will also be useful for non IEEE 802 networks and TVWS Devices. IEEE802.19.1 logical system architecture defines the system entities, which are the core of providing the coexistence for the TV band networks, and external entities which provide information to the coexistence system.


of 97

IEEE 802.19.1 coexistence system consists of three logical entities, Coexistence Enabler, Coexistence Manager, and Coexistence Discovery and Information Server. A logical entity is defined by its functional role and its interfaces with other IEEE 802.19.1 coexistence system entities, or external elements. The system may interact with two external entities, TVBD and TVWS database. They are defined outside this standard. The TAG deals with coexistence between unlicensed wireless networks. Many of the IEEE 802 wireless standards use unlicensed spectrum and hence need to address the issue of coexistence. These unlicensed wireless devices may operate in the same unlicensed frequency band in the same location. This can lead to interference between these two wireless networks. Relevance to COGEU architecture specification: Coexistence management in the Secondary spectrum market is a key feature to enabling sufficient QoS in the TVWS. As for this TAG above, COGEU architecture framework will not only consider coexistence for IEEE 802 networks and devices but will also address coexistence between non IEEE 802 networks and TVWS Devices.

2.4- IEEE 802.22

Efforts are currently ongoing to specify air-interfaces for use in white space spectrum. The IEEE 802.22 working group was formed in November 2004 as the first worldwide effort to define a novel wireless air interface (i.e. MAC and PHY) based on Cognitive Radios. The IEEE 802.22 WG is given the task to develop a CR based Wireless Regional Area Network PHY and MAC layers for use by license–exempt devices in the TV white space spectrum. Since IEEE 802.22 is required to reuse the fallow TV spectrum without causing any harmful interference to incumbents (i.e., the TV receivers), cognitive radio techniques are of primary importance in order to sense and measure the spectrum and detect the presence/absence of incumbent signals. 802.22 focus on Rural Broadband Wireless Access, a brief overview of the main interest of 802.22 are given below:

Core Technology - Cognitive radio technology based un-licensed use, primarily designed to operate in the TV White-spaces from 54-862 MHz, on a non-interfering basis with the primary users (incumbents).

Representation – Commercial industry, Broadcasters, Govt., regulators, and Academia

Projects – IEEE 802.22, IEEE 802.22.1, IEEE 802.22.1

PHY - Optimized for long channel response times and highly frequency selective fading channels.

MAC – Provides compensation for long round trip delays. The current 802.22 draft MAC employs the superframe structure. At the beginning of every superframe, the BS sends special preamble and SCH (superframe control header) through each and every TV channel (up to 3 contiguous) that can be used for communication and that is guaranteed to meet the incumbent protection requirements. During the lifetime of a superframe, multiple MAC frames are transmitted which may span multiple channels and hence can provide better system capacity, range, multipath diversity, and data rate.

Unique features introduced for Cognitive Radio based operation: spectrum sensing, spectrum management, intra-system co-existence, geo-location and security.

Mobility and Portability - Portability – IEEE 802.22 allows portability (nomadic use). In case the rules do change, IEEE 802.22 PHY is designed to support mobility of up to 114 km/hr (no hand-off is included in the current version).

TVWSDs connect in a point-to-multipoint architecture where the customer premises equipments (CPEs) are controlled by the centralized base station (BS). The BS not only controls the operational parameters of all its registered CPEs but also acts as a proxy to the database service. A CPE reports its location information to the BS and the BS uses this information to query for available TV channels from the database on behalf of the CPE. The BS manages TV channel selection for its network using an integrated cognitive brain – called the “spectrum manager.” Each of devices on the network is capable of spectrum sensing. The exact spectrum sensing capability employed is left open to implementation. However, the BS keeps track of the statistical performance of


of 97

the spectrum sensing capabilities. These statistics along with the individual sensing measurements made by the individual devices on the network can be combined to accomplish distributed sensing. Relevance to COGEU architecture specification: One key feature of the WRAN Base Stations is that they will be capable of performing a distributed sensing. This is to say that the CPEs will be sensing the spectrum and will be sending periodic reports to the BS informing it about what they sense. The BS, with the information gathered, will evaluate whether a change is necessary in the channel used, or on the contrary, if it should stay transmitting and receiving in the same one.

2.5- IEEE 802.11af (White-Fi)

FCC’s allowance of personal/portable devices in TVWS introduces another interesting standard named: IEEE 802.11af. In 2008, Google and Microsoft announced their interest in using TVWS for an enhanced type of Wi-Fi like Internet access, called Wi-Fi 2.0, Wi-Fi on steroids, or White-Fi. The idea was later formalized as a new standard called IEEE 802.11af, for which an 802.11 task group was chartered. 802.11af is expected to provide much higher speed and wider coverage than current Wi-Fi, thanks to the better propagation characteristics of the VHF/UHF bands. IEEE 802.11af can be understood as a wireless network with a CR-enabled access point (AP) and associated CR devices as end-user terminals. The CR APs operate on TVWS via spectrum sharing schemes, and the thus incurred time varying spectrum availability introduces new challenges. For example, upon appearance of PUs in a leased channel, the AP should relocate the CRs in the channel, which requires eviction control of in-service customers in case the remaining idle channels cannot accommodate all the spectrum demands. Although Wi-Fi over TVWS is still in its infancy, its resemblance to today’s Wi-Fi hotspots suggests that it may have a huge market potential in CR-based wireless networks. IEEE 802.11af is an amendment to 802.11mb/D6.0 whose implementation in solutions is likely to receive FCC approval for operation in the TV White Spaces. It follows the following main principles:

The amendment should not duplicate functionality that is being standardized in other Task Groups that are likely to complete before 802.11af.

There is no need for backwards compatibility with 2.4 GHz ISM operation.

Use the OFDM PHYs with 5-, 10- and 20-MHz channel widths to specify the basis for a system that the regulators can approve for operation in the TVWS bands.

If the FCC changes the rules, the Task Group should change the amendment accordingly. As introduced by the last bullet point above, 802.11af is a regulatory-driven amendment. Moreover, regulatory (US) rule did change recently by releasing rule FCC 10-174. The main changes in the rule can be summarised as follow:

Sensing requirements dropped o this was the biggest hurdle for low-cost TVWSDs o Sensing-only devices is now a future option

Two TV channels will be reserved nationwide for protected wireless microphones; licensed microphone category expanding to include theatres, sports arenas and churches

Incumbents needing more than two TV channels in any market for special events and sports venues can register for and reserve additional TV channels for wireless microphone use for these special events

Eventually, the 802.11af amendment has been extensively updated in 2011 taking into consideration those last rule changes. The draft is now marked as “Re-circulating”. Members of the Task Group are confident that Final 802.11 WG Approval will occur during March 2013. Given that the Wi-Fi Alliance's efforts largely parallel the IEEE's, it can be expected to see something akin to the "11n Draft 2.0" step hit the market for 802.11af before full standards ratification, given the declared hunger for product in the white space frequencies.


of 97

Relevance to COGEU architecture specification: In the case above, maybe IEEE 802.11af will be seen commercially before IEEE 802.22 due to the speed of standardization and the ecosystem inherited from the WiFi alliance, even though IEEE 802.22 have existed for many years now. COGEU consortium will be following very closely this standardisation process by keeping a privileged relationship with RIM (Research In Motion) EAB members who chair the 802.11af Task Group. While the 802.11af do not really deal with network architecture, it will be very interesting If commercial products are available before the end of the COGEU project. 802.11af could then be the perfect candidate to implement scenarios as described in section 6.2-. At first sight, IEEE 802.11af could seem that it overlaps with the already existing 802.22. Apart from the obvious pre-requisite use of a spectrum database, the differences from WRAN include:

- CSMA/CD is used. - Mobile Masters exists - Multimode devices (Client/Master) are expected - Existing definition mechanisms for bands (in .11) - Existing MAC capabilities and infrastructures for security, handoff, management, etc - Multi-band operation (e.g. TV band + 2.4 + 5 GHz in one device) common MAC over multiple bands, - Larger deployments with possibility of higher densities of master devices - Very low power devices (requirements for power conservative and efficient protocol) - Possible applications in non-TV bands (e.g. radar avoidance) - Built-in coexistence / robust signalling (based on mature CSMA/CD, MIMO, etc.)

2.6- ETSI RRS

The work is being focused on functional architectures for SDR, cognitive radio and resource optimization, SDR-based handsets and base stations, and the role of RRS in public safety and defence. Given the diversity of ways that RRS is expected to operate, it is essential to guarantee that all the systems independently standardized can cooperate in a common RRS framework. ETSI RRS is composed by four WG: WG1 – System Aspects; WG2 – Radio Equipment Architecture; WG3 – Functional Architecture and Cognitive Pilot Channel; and WG4 – Public Safety. Regarding the TV white spaces, ETSI RRS is evaluating use case scenarios for wireless and cellular technologies (e.g., LTE) and public safety. However, there is also a strong commitment in protecting primary users (or incumbent systems), namely the following:

Terrestrial Broadcasting Service (BS) including DVB-T in particular Program Making and Special Event (PMSE) services including radio microphones in particular Radio Astronomy Service (RAS) in the 608-614 MHz band Aeronautical Radio Navigation Service (ARNS) in the 645–790 MHz band

The use cases currently being proposed may be found in TR 102 907 “Use Cases for Operation in White Space Frequency Bands”. For COGEU project, the relevant use cases are:

Mid-/long range wireless access over white space frequency bands (with no/low/high mobility) Short range wireless access over white space frequency bands, namely networks with

centralized coexistence management Sporadic use of TV white space frequency bands

o Lighter infrastructure deployment through larger cell sizes o Increased spectral efficiency through reduced propagation loss o Increased spectral efficiency extended macro diversity o TVWS band-switch in case that incumbent user re-enters

At this moment COGEU contributions to ETSI are focused in providing at least one use case that is not being considered by ETSI in the document TR 102 907 and update the document. This use case is one that is being investigated in the COGEU project and is related with the spectrum market. Namely, COGEU is proposing that secondary players may purchase the spectrum with temporary exclusive rights. The exclusive rights are acquired through the central point, the broker. This would boost the investors’ confidence, since in this scenario the QoS could be guaranteed without affecting the transmissions of incumbents and being interrupted when incumbents appear.


of 97

For the future, COGEU participation on ETSI RRS will not only be focused on the proposed use case, but also in extending some already proposed use cases, namely the ones that are related with LTE over TVWS, with the novelty of spectrum trading and temporary exclusive rights. Relevance to COGEU architecture specification: ETSI RRS considers an important issue might be coordination of band allocation or utilization in the case of TVWS. Opportunistic TVWS usage involving additional eNodeBs in order to off-load traffic from congested licensed cellular bands must be accompanied by smart TVWS band allocation/sharing strategies. Since the activation and utilization of some or many bands in TVWS may change locally over time, this corresponds to a kind of changing infrastructure of active network elements in a particular band. This Scenario fits very well with current COGEU studies, and more specifically LTE-over-White-space Scenario defined in COGEU Deliverable 3.1.

2.7- CEPT/ECC SE43

The ECC WGSE (Spectrum Engineering) has set up a special project dealing with cognitive radio matters. The SE43 was set up in May 2009 and the current ECC Report 159 “Cognitive radio systems in the ‘white spaces’ in 470-790 MHz’” is still an active document. As of February, 2011, this report exists in draft format and can be downloaded from the CEPT/ECC website. The main focus of the report is, as the title suggests, on coexistence with incumbent or primary systems. It introduces definitions of “White Space”. The definition of “White Space” is taken from CEPT Report 24 “Technical considerations regarding harmonisation options for the Digital Dividend.” The report defines different scenarios for CR operation in terms of TVWSD types (personal/portable, home/office and public access points) and also discusses the three well known types of cognitive techniques: spectrum sensing, geo-location and beacons. The report is focussed on protection of four possible incumbent systems: broadcast systems (BS), Program making and special events (PMSE), radio astronomy (RAS) and aeronautical radio navigation systems (ARNS). Comprehensive data on possible sensing and separation distances are given, and ends in operational and technical characteristics for white spaces devices to operate in the band. An estimate of available white space is also included. Relevance to COGEU architecture specification COGEU considers the requirements of the draft report from SE43 of high importance. This report is intended to be the basis for a future recommendation for cognitive radio operation in Europe, a recommendation which most regulators will probably endorse.

2.8- Weightless Protocol

Current proposals for protocols to gain access to the TVWS include the Weightless protocol. The draft Weightless Specification, available for access on the http://www.weightless.org website, provides detailed information about the standard and in essence provides for the following:

Flexibility in the data rate provided depending on the application, range and environment. This includes the use of variable spreading factors which allow a trade-off of range against data rate for each terminal.

Time division duplex (TDD) operation since it may be difficult to find a pair of white space channels with appropriate duplex separation in all areas.

Relatively long frame duration of the order of 2s so that when high spreading factors are used on the frame header it remains a small percentage of the overall frame duration.

Frequency hopping at the frame rate to minimise the impact of interference - both received and caused.

A design that minimises costs and power consumption.

A broadband downlink using single carrier modulation within a 6-8MHz TV channel.

http://www.weightless.org/


of 97

A narrowband uplink with 96 uplink channels per downlink channel to accommodate the lower terminal transmit power while maintaining a balanced link budget.

Optimised MAC-level protocols that result in small headers per transmission and hence little overhead even when the payload is only a few bytes long.

Relevance to COGEU architecture specification While protocols to gain access to TVWS is of particular interest for COGEU, this protocols is not yet mature nor detailed enough to make it a base for our TVWSD-to-Database messaging interface. As stated in chapter 5.8- COGEU is following the current state of IETF PAWS WG recommendations instead.

2.9- IETF PAWS Working Group

The Protocol to Access WS database (PAWS) aims at addressing the fundamental problem of enabling a radio device to determine, in a specific location and at specific time, if any white space is available for secondary use. More specifically, the overall goals of this working group are to:

1. Standardize a mechanism for discovering a white space database. 2. Standardize a method for accessing a white space database. 3. Standardize query and response formats to be carried over the database access method. 4. Ensure that the discovery mechanism, database access method, and query/response formats

have appropriate security levels in place. In completing its work, the group will:

Evaluate existing discovery mechanisms to determine if one of them provides the necessary application features and security properties (or can be extended to do so) for discovering a white space database. Examples might include DNS.

Evaluate existing application-layer transport protocols to determine if one of them provides the necessary application features and security properties (or can be extended to do so) for use as a building block for communication between location- aware devices and white space databases. If such a method exists, the group will reuse it; if not, the group will develop one. Examples might include HTTP.

Develop a method for querying a white space database. Such a method will utilize, so far as possible, the features of the application-layer transport protocol and not re-implement them in the new protocol. It will include mechanisms to verify that the requests and responses come from authorized sources, and that they have not been modified in transit. Examples might include LDAP.

Define extensible formats for both location-specific queries and location-specific responses for interaction with radio white space databases. The group will consider whether existing data formats can be reused.

The protocol must protect both the channel enablement process and the privacy of users. Robust privacy and security mechanisms are needed to prevent: device identity spoofing, modification of device requests, modification of channel enablement information, impersonation of registered database services, and unauthorized disclosure of a device's location. The group will consider whether existing privacy and security mechanisms can be reused. Relevance to COGEU architecture specification Even if the work of PAWS is still experimental and only Internet Draft has been released yet, the COGEU protocols for communication between players and the COGEU broker has been done while following PAWS WG advances. A close attention has been given to PAWS work related to the method for accessing a white space database, and the query/response formats for interacting with a white space database. COGEU won’t be able to use PAWS work in order to define the structure and contents of the databases themselves because this task is out of scope of the PAWS working group.

2.10- ICT-QOSMOS

QoSMOS is a Framework 7 Integrating Project from call 4. The project began on 1st January 2010. Co-ordinated by British Telecommunications plc in the UK, it has 14 partners from across Europe and one


of 97

from Japan. The primary objective of QoSMOS is to develop a framework for Cognitive Radio systems and to develop and prove critical technologies using a test-bed. Involvement in standards bodies and industrial forums is emphasised from the start, to increase the probability of adoption of QoSMOS results into standardised products. The initial focus is on opportunistic use of radio spectrum, with an early example being TV White Spaces. Figure 3 gives an illustration of the QoSMOS concept. The QoSMOS concept includes the use of two cognitive managers (CM), which operate on different timescales and amounts of radio resource. The lower one in the figure is centralised and operates on a longer timescale, it builds a portfolio of the available resource in a particular region. This CM manages a set of rules, whose parameters can be initially programmed for different regulatory regimes. The upper CM is distributed and operates on a shorter timescale, allocating spectrum to individual wireless streams from the portfolio, also to a set of rules. Feedback flows from the upper one to the lower one to adjust and optimise the rules. In order to make decisions about spectrum occupancy, equipment is required to have highly sophisticated sensing mechanisms and a system of metrics which enable them to correctly detect temporarily and/or spatially unused spectrum (the so-called spectrum holes) and make decisions on whether to use these without causing harmful interference to other users. This ability to make decisions to occupy spectrum is what is meant by a ‘cognitive’ approach. Use of the spectrum will become more efficient if local decisions are made on what spectrum to occupy for a particular service at a particular time. If a mix of distributed triggers (e.g. from the user terminals) and centralised triggers (e.g. from the network) can be used in a cognitive decision making process, cost savings will be achieved through massive reuse of spectrum resources without the burden of large upfront investments.

Figure 3: Illustration of the QoSMOS concept Co-existence with and protection of other services sharing the spectrum is of utmost importance and is a differentiator between the US and Europe, issues continue to arise with European regulators taking defensive action against incompatibility of US 802.x equipment. For example there is activity in the US to pave the way for marketing of 802.22 TVWS equipment in Europe. This is not desirable for the protection of EU services that share the TVWS bands because of the population map, the geographical topology and the borders between countries (and hence different regulatory regimes). In QoSMOS, a framework is provided that allows policy adjustments to a spectrum manager to suit different EU countries (and the US).


of 97

Relevance to COGEU architecture specification Although the problem setting for QoSMOS and COGEU has some correlation, it will be shown that the COGEU approach extends the current commons spectrum usage by introducing secondary spectrum trading based on a centralized entity, namely the broker, to enable co-ordinated usage of the TV white spaces.

2.11- ICT-SACRA

In the SACRA project, the technical approach consists first in the definition of a target scenario for the study, in the specification of global system requirements (architecture, target figures, characteristics) and in the definition of working assumptions, parameters and hypothesis (WP1). Based on this common framework, the different enabling techniques are studied and beyond state-of-the-art solutions are proposed. WP2 addresses the sensing and access techniques, and especially advanced space-time frequency polarization coding schemes. WP3 is dedicated to the radio resource management and provides interference management and allocation techniques for multi-band operation. WP4 addresses the design of antenna and radio frequency parts: integrated RF receiver front-end and versatile ADC, compact multi-band dual polarized multiple antennas, architecture for an integrated RF transmitter, DAC and power amplifier pre-distortion. WP5 will address the flexible base band design by providing a framework for embedded software design and validation. All these studies will allow to finally form a compound system integrated in a single platform, to be validated and tested in the scope of WP6. In WP1, the system specifications will be refined along the project, taking into account the results achieved in the enabling techniques related studies. WP1 will finally provide a recommended system definition, with associated techno-economical study. Dissemination of the project results, especially through workshops and proposals to standardization is addressed in WP7.

Figure 4 : SACRA ICT technical approach SACRA and COGEU complementary approach SACRA and COGEU are good examples of consistency at the European research level. The two projects are complementary in their approach. SACRA provides design and development of a cognitive and multiband terminals (in licensed and TVWS bands), whereas COGEU focuses on the definition and implementation of a framework for TVWS usage relying on SDR existing platforms.


of 97

2.12- Cross analysis for COGEU system architecture

To our knowledge, there is no current architecture proposed by standardization bodies for secondary spectrum trading, that is, all of them assume unlicensed access (spectrum commons regime). Spectrum commons regimes promote sharing, but do not provide adequate quality of service (QoS) for some applications. COGEU consortium is convinced that unlicensed use of TVWS bands is not fully adequate solution for all possible applications which may apply in Europe. Therefore COGEU endeavour to promote the combination of spectrum commons regimes and temporally exclusive rights for use within Europe. The State-of-the-art above provides an analysis of various IEEE bodies, ETSI RRS, IETF WG, and ICT projects which are relevant to COGEU. Some of them have a high-level scope, addressing system architecture issues, while others address more specific aspects like radio access techniques, sensing, and incumbent protection. Moreover, the cross-analysis between those technologies and the requirements coming from WP2 show that a lot of concepts can be re-used, e.g., the Dynamic Spectrum Manager defined in ETSI RRS or the Centralized spectrum management concept introduced in 802.22 (see Figure 5) as a baseline for the COGEU reference model. Distributed sensing and measurements concept specified in IEEE 802.22.1 is also considered for one of the possible regulatory scenario (see section 3.2). The following is an attempt to highlight the high diversity of database proposals. The following figures obviously emphasize that no standardized architecture can fit all the specific needs of COGEU system requirements, if taken as it is. For example, main achievements of 802.22 Wireless Regional Area Network related to COGEU architecture analyses are:

802.22 has established spectrum sharing techniques that can be readily adapted to enable coexistence with other 802 technologies

802.22 spectrum manager is capable of autonomously deciding which coexistence mechanism to use and how

802.22 has developed a basic interface for incumbent database service

802.22 will consider providing interface between its spectrum manager and 802.19 coexistence manager

Figure 5 shows an overview of interfaces provided by 802.22 spectrum manager.


of 97

Figure 5: Spectrum manager interfaces ( Source: [21])

A second example highlighting current database diversity is the shown in Figure 6 from The White Space Database Group, created by Google while introducing their geolocation database proposal. It has been useful for COGEU reference model definition, in order to identify the specific need for regulatory enforcement interface.

Figure 6: Unlicensed Operation in the TV Broadcast Bands (Source: [20]) While there is no need in COGEU for CPC (Cognitive Pilot Channel) to support the user terminal for discovery of available radio access and reconfiguration, another architecture consideration relevant to COGEU is ETSI RRS Database access proposal seen in Figure 7. The CPC is defined as a channel that conveys the elements of necessary information facilitating the operations of the Cognitive Radio Systems. The CPC provides information on which radio accesses can be expected in a certain geographical area. This information includes operator information, RAT type as well as used frequencies. In COGEU we consider our centralised database/broker based allocation slightly different than the CPC approach since TVWSD are not necessary very complex Cognitive Radio Systems.


of 97

Figure 7: ETSI RRS Database Access (Source: [19]) Finally, considerations as expressed in 802.19.1 (see Figure 8) are of particular interest in dealing with coexistence. In order to insure coexistence, the 802.19.1 architecture specification draft defines:

Coexistence Enabler o Request and obtain information, required for coexistence, from TVBD network or device o Translate reconfiguration requests/commands to TVBD-specific

Coexistence Manager o Discovery of other CMs o Decision Making o Support exchange of information

Coexistence Discovery and Information Server o Support Discovery of CMs o Collect, aggregate information


of 97

Figure 8: IEEE 802.19 co-existence framework (Source: [22]) Moreover, an architecture mapping is defined between 802.19.1 and 802.21 that could be of particular interest for handover preparation in the scope of mobility management studies in COGEU. In conclusion, one of the main statements made through this analysis concerns the diversity on databases proposal for spectrum management. Indeed, standards and technology proposals presented above have different visions on what a spectrum database should be, depending on the context and issues considered. Nevertheless, there is a clear need to adapt the state of the art on architecture proposals in order to fully satisfy COGEU requirements. The rest of this deliverable will provide analysis about which type of adaptations are needed, and present the final COGEU Architecture.


of 97

3- COGEU approach and regulatory scenarios

COGEU investigates both the commons and secondary spectrum trading approaches in the TVWS. In a commons spectrum usage model, there is no spectrum manager to preside over the resource allocation. Spectrum commons regimes promote sharing, but do not provide adequate QoS for some applications. However, for applications that require sporadic access to spectrum and for which QoS guarantees are important, temporary licensed spectrum with real-time secondary markets may be the best solution. Trading allows players to directly trade spectrum usage rights, thereby establishing a secondary market for spectrum leasing and spectrum auction. Of course, both regimes, spectrum commons and spectrum trading, are only possible to the extent allowed by national regulation. For that matter, COGEU assumes two regulatory scenarios as follows:

1. Both geo-location database access and spectrum sensing are required for the protection of Incumbents. 2. Only geo-location database access is required for the protection of the Incumbents.

Before giving the details of these two scenarios, we first present an overview of the COGEU system requirements from previous deliverables.

3.1- Overview of COGEU system requirements

The following is the list of general COGEU system requirements identified in previous deliverables with impact on COGEU architecture:

COGEU will follow CEPT definition of TVWS as a part of the spectrum, which is available for a radio-communication application (service, system) at a given time in a given geographical area on a non-interfering non-protected basis with regard to primary services and other services with a higher priority on a national basis. [D2.1]

COGEU operation band starts from channel 40 to channel 60 (622 MHz-790MHz). Efficient SDR platform, RF modules and compact antennas are required to operate in this band. [D3.1]

The system shall provide means to protect incumbent systems. COGEU approach is combining the use of geo-location database together with autonomous sensing. [D2.1]

The system should provide a signaling channel for reporting of local sensing data and supports centralized cooperative sensing. [D2.1] [D3.1]

The database protects DVB-T and professional PMSE systems that can be planned in advance. Other PMSE users such as ENG (Electronic News Gathering) shall be protected trough autonomous sensing. [D3.1]

The maximum allowed power for TVWS operation is not pre-fixed. The power level is limited by the geo-location database information for protection of incumbent systems. Power control shall be implemented in COGEU transceivers. The maximum allowed transmit power in a specific vacant DVB channel is computed based on co-channel and adjacent channel protection ratios. [D3.1]

COGEU database will provide the “validity period of information” i.e. a period after which a database query should be repeated. This would allow for flexibility and minimization of the overhead if, for instance, no PMSE users are at the specific location. Note that if time validity is provided then a general update frequency is not needed (2h in initial OFCOM proposal). [D4.1]

Cross-border issues have to be considered in the specification of the database. [D4.1].

COGEU devices acquire white space interfaces alongside other more established radio interfaces. COGEU envisage that the initial access to the geo-location database by unlicensed WSD will use existing radio interfaces such as WiFi, LTE or WiMax. [D4.1]

The system shall be able to facilitate coexistence with other secondary systems operating in TVWS. This is done through dynamic TVWS allocation mechanisms based on protection rules specified for each combination (e.g. LTE over TVWS vs. WiMAX over TVWS). [D3.1]

The system should provide a means to support applications with different QoS parameters, such as transmission rate, delay and delay jitter. Thus, COGEU system should be flexible enough to satisfy different QoS requirements. Because TVWS devices have to do spectrum sensing, database access and reconfiguration before transmitting, some delay sensitive


of 97

services, such as voice service, should be considered when designing the RRM strategies (WP6). [D3.1]

The COGEU WSDs\Broker are prohibited from make use of TVWS until they have successfully determined from the database which frequencies, if any, they are able to transmit on in their location. [D4.1]

For the COGEU WSDs (spectrum commons) a master-slave configuration is envisage, where the master connects to the database and the slaves are managed by the master, without access to the database [D4.1].

In line with the trading mechanism, price discovery is an important requirement to enable the Broker to allocate the spectrum to the most valuable players. Efficient spectrum policies to enable fairness in the spectrum sharing models needs to be implemented. COGEU considers a regulatory regime that allows a fast re-assignment of spectrum ownership. [D3.1]

The centralized point in the COGEU system allows, in the event of an emergency, a higher priority policy be propagated into the spectrum broker temporarily rescinding non-emergency utilization of the TVWS in the specific areas of need. Moreover, service prioritization can be incorporated into the database. Public Safety systems would have the highest priority. [D3.1]

For regulatory enforcement, the database can be used in identifying the source of harmful interference where it occurs and may enable a “remote de-activation” of the device. [D3.1]

The system should have a web interface able to allow players to follow the secondary market activity, communicate with the broker and negotiate spectrum rights in real time. [D3.1]

COGEU will adopt Internet-based protocols and standard enquiry languages. The proposed database access procedure includes XML through web services. [D4.1]

The COGEU system should be able take advantage of regulatory changes in incumbent protection requirements, like the reduction or elimination of autonomous sensing requirements. [D3.1]

COGEU must not block the evolution of incumbent systems such as DVB-T2. [D3.1] The following table summarise the system requirements (requirements appear here, based on their assigned priority) while providing an overview of the features currently being implemented under the scope of WP7. Crosses in this table allow to identify which module provides a given feature. Only requirements related to the architecture definition are listed in this table.

On-going implementation in

WP7

System Requirements Ge

o-l

oca

tion

Da

tab

ase

we

b

se

rvic

es

TV

WS

occu

pa

ncy r

ep

osito

ry

Bro

ke

r In

pu

t m

od

ule

Dyn

am

ic R

ad

io E

ng

ine

Coe

xis

tence

em

ula

tor

mo

du

le

Po

licie

s R

ep

osito

ry

Po

licie

s R

ep

osito

ry W

eb

se

rvic

es

Sp

ectr

um

Allo

ca

tio

n p

roce

ss

Pe

rfo

rma

nce

Mo

nito

r

Ma

ste

r W

SD

in

terf

ace

Pla

ye

r com

mu

nic

atio

n M

odu

le

Incumbent protection

X X X X

Coexistence X X X X X X X

QoS requirements of SU

X X X X X X X

Geo-location DB access X

X X X

Secondary Spectrum Trading

X X

X X X

Regulatory enforcement/changes

X X

Location Information provision

X X


of 97

Service Prioritization X

Frequency Planning

X

Autonomous sensing

Mobility

The following sections present the assumed regulatory scenarios under which these requirements will be fulfilled accordingly.

3.2- Regulatory scenario 1: Spectrum commons with sensing and geo-location access required

Information on DVB-T incumbents is stable and hence suitable for the spectrum database approach. The same is the case with registered PMSEs, usually for professional applications. COGEU assumes that a database for professional PMSE is either available or will be built up in advance of introduction of white space using equipment. However, the unpredictability of unregistered PMSE applications and Electronic News Gathering, which requires protection, is the main challenge in the design of the COGEU geo-location database. Moreover, so far there is no clear regulatory framework regarding sensing requirements in Europe. Because the process of switching PMSE to “safe harbor” will take years to be concluded in Europe, therefore we can assume a scenario where TVWS commons has to coexist with unpredictable PMSE trough combination of cooperative sensing and geo-location database access (master-slave topology) should be assumed for unlicensed use of TVWS. The implications of the 1

st Regulatory Scenario are as follows:

In order to provide means to protect incumbent systems, combining the use of geo-location database together with autonomous sensing seems plausible. The database protects DVB-T and professional PMSE systems that can be planned in advance. The maximum allowed transmit power in a specific vacant DVB channel is computed based on co-channel and adjacent channel protection ratios. Other PMSE users (not planned, not registered) such as ENG shall be protected trough autonomous sensing.

In this scenario, autonomous sensing should be mandatory for PMSE and optional for DVB-T signals (which are mainly protected by the geo-location database). Detection thresholds are adopted from current regulatory framework. The system should provide a signaling channel for reporting of local sensing data and supports centralized cooperative sensing.

Combining the two approaches can relax the sensitivity required for sensing devices which is a major limitation of TVWS developments. Also, since local sensing is only performed in a limited number of TV channels indicated by the database, the hybrid approach will speed up the sensing process. Moreover, cooperative sensing exploits spatial diversity of sensors located in different positions (preferentially with low correlated shadowing). COGEU will investigate cooperative sensing able to relax the sensitivity requirements of single nodes trough the reduction of the hidden terminal margin. Cooperative sensing requires protocols for sharing sensing information among TVWS devices which add extra complexity and sensing overhead to the TVWS system.

The system should provide a means to support applications with different QoS parameters, such as transmission rate, delay and delay jitter. Thus, COGEU system should be flexible enough to satisfy different QoS requirements. Because TVWS devices have to do database access (and spectrum sensing, if necessary) and reconfiguration before transmitting, some delay sensitive services, such as voice service, should be considered when designing the RRM strategies.


of 97

Figure 9 shows the initial COGEU frame structure for the combination of geo-location data base access and sensing.

DetectionDatabase

Access

Data

TransmissionDetection

Data

Transmission... Detection

Database

Access

Data

Transmission...

Minimum 2h

60 s

Up to 2 s

Sensing timeReporting

to Master

Cooperative

decison

Reporting

to CR

Time to leave the band

or reconfiguration time

Detection

Figure 9: Initial COGEU frame structure [Source: D3.1].

However, wireless service applications in this scenario will have to deal low QoS (latency). Therefore, there is a need for a scenario where better QoS guarantee can be given. In the next section we present a scenario where sensing is not required, only geo-location database access. Detailed discussion of this scenario including the reference model will be given in chapter 4-.

3.3- Regulatory scenario 2: Commons and secondary trading, only geo-location access required

For reliable access to the TV white spaces and the guarantee of the QoS for wireless service providers, COGEU envisions a scenario where geo-location database access and “safe harbor” channels for non-registered PMSE will be required. Within this scenario we assume that sensing is not necessary. The proposed solution is to consider that Europe has implemented “safe harbor” for the exclusive PMSE usage, i.e., number of TVWS channels for reserved PMSE usage only in which no TVWS devices would be permitted. The “safe harbor” bands are flexible and it may change from country to country. These channels are excluded by the geo-location database and therefore out of the market. In this case the broker doesn’t need to consider sensing (only database access) and system doesn’t need backup channels to guarantee QoS and increasing spectrum efficiency. Hence, COGEU considers a regulatory regime that allows a fast re-assignment of spectrum ownership. The rationale for this scenario is aligned with CEPT ECC Report 159:

“… it appears that the identification by national administration of at least one (or more) safe harbor channel, not used by DTT and which would be reserved for PMSE use would be helpful for the protection of PMSE, in particular for casual or unplanned usage by PMSE which would not be registered.”

Therefore, COGEU considers this regulatory scenario assuming that the TVWS in channels 21-40 are reserved for PMSE use (here just non-predictable ENG use is relevant as the other predictable systems can be registered in a database). This is in line with COGEU assumptions where only channels 40-69 are considered. This gives a stable situation for COGEU considerations. Reserving some channels (e.g. the available TVWS in channels below 40) for ENG would stabilize the available TVWS in the (other) channels considered by COGEU. These TVWS are for spectrum commons use and for spectrum broker as will be discussed later in the deliverable. If there arises a situation where incumbents need more spectrum, then the spectrum available for spectrum commons and spectrum broker will be reduced by the regulator or its representative (i.e. also here the situation of less TVWS may happen). The broker will have to cope with this anticipated situation. The safe harbor concept smoothes variation of TVWS availability and, to some extent, has impact on the cost of available TVWS for secondary use. The benefit for TVWS secondary use is that it is known which channels are reserved.


of 97

Currently, only unlicensed access to TVWS is envisage\allowed by regulators, typically for low power applications (CEPT, OFCOM, FCC). COGEU investigates an extension of this regulatory regime and proposes a secondary spectrum market of TVWS that can leverage the value of these underutilized bands. The implications of the 2

nd Regulatory Scenario are as follows:

The system shall be able to facilitate coexistence with other secondary systems operating in TVWS. This is done through dynamic TVWS allocation mechanisms based on protection rules specified for each combination.

Cross-border issues have to be considered in the specification of the database. In COGEU context, assisted-GPS technology will be adopted in order to overcome problems with low signal levels.

Players (spectrum holders and spectrum seekers) are subject to regulation. Regulators must determine exactly what rights can be granted to secondary users. A centralized spectrum broker manages TVWS on real-time basis subject to non interference rules. Moreover, specification of spectrum usage rights and obligations, the minimum set of information that parties to a spectrum trade must disclose, and approaches to the protection of competition needs to be investigated. It is important to note that this model has the potential not just to open the market to new players but that it also has the potential to create new business opportunities for the spectrum broker entity be it in new public sector roles or in the commercial sector.

Detailed discussion of scenario 2 will be given in Chapter 5-. Furthermore, the following core elements for an efficient secondary spectrum market should be taken into account in the subsequent development of the project [2][3]:

a large number of buyers and sellers to create competition necessary for an efficient market,

clearly defined rights to the spectrum for both buyers and sellers,

free entry and exit to the secondary markets,

availability of relevant information to all buyers and sellers,

a mechanism to bring buyers and seller together and facilitate the transaction with reasonable administrative costs and time delay,

reliable procedures for payment between players, etc. These features will be partly addressed in the COGEU architecture and in the subsequent work packages.


of 97

4- Architecture for spectrum commons with sensing and geo-location access required

The main objective of this chapter is to derive reference architecture for the first COGEU regulatory scenario considered under the 1

st Regulatory Scenario. For the COGEU TVWSDs (spectrum commons)

a master-slave configuration is envisage, where the master connects to the database and the slaves are managed by the master, without necessary access to the database. The master-slave architecture is presented in section 4.1-.

Figure 10 : Reference Model for spectrum commons with sensing and geo-location access required

The geo-location database has a central role in the general architecture of COGEU. In COGEU model, the regulatory bodies assign TVWS for spectrum commons (free access) in given areas. In the current regulatory assumption, the COGEU geo-location database receives enquires from TVWSD as illustrated in Figure 10. The main purpose of the COGEU geo-location database is to enable the protection of the incumbent systems from harmful interference (DVB-T and PMSE). Besides the information on the incumbent systems that a database will hold, it will also include the geo-location information per geographic pixel for a specific region and records of the WSD that operate in the specific region. The structure is designed to accommodate data about the incumbent systems. The information must come from a verified source (such as the national regulator) since is the base of calculation of all available TVWS bands. Furthermore, the database include regulatory information, information from a central database regarding the border incumbent stations, and a record of the white space devices that access or can access the database. An important part of the database is the geographical data. These data holds the terrain and building information available in order to calculate the available TVWS. In order to implement a consistent demonstrator, partners from WP7 have proposed a data structure for the geolocation database illustrated in Figure 11.


of 97

Figure 11 : Geo-Location Database Data Structure

As reported in D4.1, the COGEU geo-location database is accessible by the following interfaces as shown in Figure 12

Interface A is to provide communication with the WSD repository that operates under the spectrum commons operations;

Interface C is connected to a regulation and policies repository for the current area that the database is operating;

Interface D will be by the Incumbent systems repository which will provide information for the protected incumbent systems;

Interface E is public access interface that would enable anyone to search the Database’s non-confidential publicly available information.

Interface F connects the local database with the central database in order to retrieve updates on policies and information regarding close border areas.

Each of interfaces described above will use IP security for the obvious reasons.


of 97

Figure 12: COGEU interfaces with the geo-location database [Source: D4.1]

4.1- Master-slave architecture

In line with current regulatory bodies (FCC, OFCOM) device distinction, COGEU considers two types of TVWS Devices:

“master devices” that contact a database to obtain a set of available frequencies in their area; and

“slave devices” which obtain the relevant information from master devices but do not contact the database themselves.

For example, a master device might be a wireless router in a home – similar to a WiFi router but operating in the interleaved frequencies – while the slave devices might be other wireless devices in the home such as laptops and printers that are connected to the router. In order to be used legally within, any device that emits a radio signal must either have a licence or be exempted from licensing. In the Spectrum Commons regulatory scenario considered, it is proposed to exempt TVWSDs from the need to be licensed.

4.1.1- Master devices

Master devices are generally fixed location devices communicating with the central geolocation database. In line with OFCOM recommendations, exemption from licensing of a “master device” would be subject to the following:

i) Determining its location and assessing the accuracy of that location with 95% certainty. This location accuracy should reflect the maximum area of operation within which slave devices can be located. ii) Consulting a list of geolocation databases and selecting one of these databases unless it has previously consulted the list within the last 24 hours. iii) Sending its location and accuracy of that location to the selected geolocation database along with its model identification and for devices mounted on a master or similarly its height above ground level. iv) Receiving from that database a set of parameters including the frequencies of allowed operation, associated power levels, geographic validity of operation and time validity of operation. Other parameters may also be provided. v) Operating in accordance with these parameters, ceasing transmission immediately where the time validity expires or where it moves outside of the geographic area of validity.


of 97

vi) Operating in a “fair” manner, sharing the available spectrum resource as evenly as possible across competing users. vii) Provision from the device manufacturer as to the out-of-band performance of their device in terms of relative power levels emitted into adjacent bands up until n+/-9 or until power falls to the noise floor. viii) Managing slave devices as required by signaling to them the parameters by which they may communicate with the master device. ix) Maintaining a record of all active slave devices and requiring slave devices to stop transmission if the master device needs to cease transmission for any reason including an expiry of time validity or moving outside the geographical area of validity.

4.1.2- Slave devices

Slave devices must receive regular signals from master devices that provide an updated list of good-to-go channels, or they must contact the master device themselves at least once per minute. Meanwhile the master device and fixed devices must check their own locations at the same rate, except if in "sleep mode"—that is, the machines aren't transmitting data, but they aren't powered down either. Exemption from licensing of a “slave device” would be subject to the following:

i) Receiving a signal from a master device indicating that a channel is available for use along with an allowed power level. ii) Operating in accordance with the signaling from the master device. iii) Ceasing transmission immediately when instructed by the master device or within 5 seconds of not receiving a response from the master device to a transmission. iv) Transmitting only to a master device (and not directly to other slave devices). v) Provision from the device manufacturer as to the out-of-band performance of their device in terms of relative power levels emitted into adjacent bands up until n+/-9 or until power falls to the noise floor.

4.1.3- Information exchange between geo-location database and WSD

As first requirements for TVWS Devices (in this case Master devices) information exchange with a geolocation database, we’ll stick to regulatory bodies, and in particular OFCOM that provides recommendations on that topic. Information Provision according to OFCOM in [23]:

Location: The device would provide its location and the accuracy of the location. Download only frequency availability relevant to its current location.

Device type: Providing information about the type of the device, it might allow information to be returned according to device capabilities.

Information returned to the device

Cognitive devices using geolocation will be prohibited from transmitting until they have successfully communicated with the database and determined which frequencies, if any, are available in their location.

Frequency availability

It seems appropriate for the database to perform the computations needed to translate the known transmitted location into frequency availability. Making these calculations to the database ensures that they can be carefully verified and changes made if necessary.

Form of Information


of 97

The simplest form of information would provide a list of frequencies that could be used within each pixel.

In order to allow variable size bandwidths to be used using start and end frequencies seems more appropriate.

In addition the maximum transmit power can be provided for each frequency assignment. This would allow the devices to operate accordingly in order to minimise the possible interference or to increase the flexibility of the device.

Considerations on Location

All the terrain covered by a geo-location database is represented as “pixels” which are squares of prearranged dimensions.

Each pixel is associated with a list of available frequencies and other relevant data that are provided to cognitive devices querying the database.

It should be noted that the area associated with the “location” as determined by a WSD may cover one or more pixels, depending on the location accuracy of the device.

The following technical information needs to be communicated by the WSD to the geo-location database:

Location

Location accuracy

Expected area of operation (optional) – coverage area

Device type

In the case of a master/slave WSD configuration, the above information will be obtained by the WSD master by requesting it from its associated slaves or deriving it by other reliable means.

The following technical information will be communicated by the geo-location database to the TVWSD:

Available frequencies (minimum requirement)

Maximum transmit power

The appropriate national/regional database to consult Also, on a more generic basis, systems operating under TVWS should follow those rules:

The TVWSD shall only transmit in the 470-790 MHz band once it has successfully communicated with the database and received the instructions regarding the frequencies available in its location and the allowed power levels.

In the case of a master/slave WSD configuration, it can be envisaged that the master would be responsible for the query of the database and that associated slaves would be controlled by the master and would receive information on their operational parameters (channels, powers, etc) directly from the master without querying the database themselves.

While technical condition of operations are yet to be finally defined by CEPT, this overview of required information exchange from OFCOM is likely to be very similar in Europe to what is detailed above.

4.2- Cooperative Sensing Architectures

In COGEU we consider a sensing architecture where a central point (Master Device) is responsible for gathering any sensing information available from the slaves WSD. COGEU sensing architecture emerged from analyzing proposals made for cooperatives sensing architecture that benefits from the geo-distribution of sensing capable devices to gather information on incumbent users.

4.2.1- Protection of incumbent users

With spectrum sensing, devices try to detect the presence of incumbent users in each of the potentially available channels. Spectrum sensing essentially involves conducting a signal strength measurement within a candidate channel, to determine whether any incumbent user is present. When a channel is determined to be vacant, sensing is typically applied to adjacent channels to determine what constraints there might be on transmission power, if any. If power restrictions cannot be filled, channels are excluded. When considering the sensing alone, a significant advantage is that it does not rely on any existing local infrastructure, such as connection to a database or a beacon, for example. This may be more important in remote and rural areas. However, if sensing thresholds are set very low, then increasing device cost and complexity is matched with a reduced number of available channels. This


of 97

would reduce the potential value to end users, particularly in areas of higher population density, and would hinder cost reductions in the technology. So far analyses of sensing performance assume that detection is carried out independently by each device, in ignorance of results found by other cognitive devices in the same location. With cooperative sensing in which devices share their findings brings the potential to reduce the hidden node margin by a substantial amount. Key parameters for spectrum sensing include: sensing threshold; periodicity of re-sensing on channels that have been detected as vacant; and sampling duration. The geo-location database appears as an alternative to the sensing. In this process devices access to a database to determine which channels are available on their current location. The restrictions, if any, may also be subject of inquiry to database. On this case, location accuracy and frequency on the database inquiry are essential. This approach would require fast update (real-time in principle needed) in order to cover/model dynamic components (time variability) in the licensed network. Beacons, that will not treated here, is also another method to determine whether a channel is vacant. The beacons are signals broadcasted in the network that may indicate that particular channels are in use by incumbent services.

4.2.2- Cooperative Sensing in Cognitive Radios

Nodes can cooperate to achieve multiuser diversity and mitigate the Hidden Node Problem, which occurs when the cognitive radio (CR) is shadowed or in severe multipath fading as shown in Figure 13.

Figure 13 : Cooperative Spectrum Sensing in CR Networks CR 1 is shadowed over the reporting channel and CR 3 is shadowed over the sensing channel.

Cooperative Sensing is performed as follows:

Every CR (slave) performs its own local spectrum sensing measurements independently and then makes a binary decision on whether the Primary user (PU) is present or not.

All of the CRs forward their decisions to a common receiver.

The common receiver fuses the CR decisions and makes a final decision to infer the absence or presence of the PU.

Advantages

Addressing the Hidden Node Problem

Increased ability to deal with worsening SNR

Increased probability of detection'

Users can employ less sensitive detectors (as the missed detection and false alarm probabilities are reduced)


of 97

As an example, IEEE 802.22 is a standard to describe a physical and media access layer of a wireless regional area network that is intended to make use, on a non-interfering basis, of unused TV broadcast channels.

4.2.3- Cluster Based sensing Architecture

In cognitive radio networks, the limitation of control channel bandwidth is a challenge of cooperative spectrum sensing when the number of cognitive users becomes very large. Cluster-based architecture is applied for cooperative sensing to avoid the congestion on control channel and reduce the sensing delay. In [26], authors proposed cluster-based cooperative spectrum sensing scheme to improve the efficiency of the network. The number of clusters effects both the system efficiency and detection performance significantly. By balancing the tradeoff between the communications overhead and sensing reliability, one can obtain the optimal number of clusters, which can minimize the cooperation overhead without any performance loss of reliability. Moreover, a clustering strategy is proposed based on a given number of clusters and simulation results show the superiority of the proposed strategy.

4.2.4- IEEE 802.22 Sensing overview

IEEE 802.22 provides TV and wireless microphone protection using spectrum sensing, as well as wireless microphone protection using beacon. While those two approaches could seems. The following is a brief overview of the functionalities provided. For TV and Wireless Microphone Protection Using Spectrum Sensing FCC R&O requires:

DTV protection at -114 dBm in 6 MHz of bandwidth. This amounts to an SNR of -19 dB for equivalent receiver noise figure of 11 dB and 22 dB safety margin at edge of coverage

Wireless microphone protection at -114 dBm in 200 kHz bandwidth. This amounts to an SNR of -3 dB for equivalent. receiver noise figure of 11 dB.

Several blind and signal specific feature-based sensing schemes have been proposed and thoroughly evaluated using TV Broadcaster supplied over-the-air collected signals

Spectral correlation based sensing, Time domain cyclostationarity, Eigen value based sensing, FFT – based pilot sensing,

Higher order statistics based sensing For wireless Microphone Protection Using Beacon:

802.22 has designed a beacon signal which will be transmitted from wireless microphone base stations with 250 mW (as compared to 10 mW for microphones). These beacon signals consist of repeated pseudo-noise (PN) sequences and occupy a bandwidth of 78 kHz.

Security features are provided for beacon authentication

4.2.5- Architecture for combination of geo-location database with sensing

Although the recent moves from FCC and OFCOM have removed sensing requirements, COGEU has to accommodate some level of local sensing because:

a) The protection of no registered wireless microphones (PMSE) is required. b) The process of moving PMSEs to specific channels and store them in a database will take

years to be finished. c) Few EU countries have a database for wireless microphones. d) PMSE usually operate indoors, implying that they are hard to geo-locate.

TVWS Devices (TVWSD) have in principle two methods to determine if a channel is occupied or not: (1) sensing, where the channel is probed to find incumbent signals at or above certain signal strength; (2) consult a geolocation database where for a certain region the attribution of the channels to primary users is presented. Both methods have their advantages and disadvantages. For sensing, to avoid hidden-node problem, the channel must be sensed at very low levels of signal strength, e.g. -126 dBm for wireless microphones, which is less than the noise floor (-121 dBm). Sensing at this level may raise the false


of 97

alarm rate. However, sensing makes devices truly cognitive and it is useful to detect signals that are not registered in databases. The geo-location database is based on predictable models to determine the coverage in a certain region, namely its boundaries, which can be quite inaccurate. Furthermore, the device needs to provide its location, e.g. by means of GPS, and access to the database (or use a preloaded device), which may be costly. However, the channel occupancy is based on databases and is independent of channel characteristics, which can greatly reduce the false alarm rate. Databases need to be updated with correct information as time goes by, otherwise errors in the database will lead to interference in the field. In this sense, a solution that combines sensing and geolocation databases seems to be the most reasonable solution, optimising channels occupancy and minimising the false alarm rate.

4.2.6- Methodology for combination sensing with geo-location database

In the first COGEU regulatory scenario TVWSD are capable of performing sensing and geo-location database access. The geolocation spectrum database is used to determine if the device is inside or outside a TVWS area. Let T be the threshold that the sensor uses, e.g. T = -120 dBm for DVB signals. Let S(T) be the decision based on the threshold T, and G be the decision based on the geo-location database, as presented in Table 1.

Table 1: Channel availability based on geolocation-only information

G = 0 The device is inside the TVWS map and the channel is vacant

G = 1 The device is outside the TVWS map and the channel is occupied

Following the same approach, for S(T) the information may be presented as in Table 2.

Table 2: Channel availability based on sensing-only information

S(T) = 0 The signal is less than T and channel is vacant

S(T) = 1 The signal is higher that T and the channel is occupied

When combining both sensing and geo-location information four situations may occur, as shown in Table 3.

Table 3: Channel availability based on geo-location and sensing information

G AND S(T) Decision

G = 0 S(T) = 0 Channel is vacant

G = 1 S(T) = 1 Channel is occupied

G = 0 S(T) = 1 The sensing information is used for the decision as shown in 1)

G = 1 S(T) = 0 The sensing information is used for the decision as shown in 2)

1) In this situation the database indicates the device is inside the TVWS map and the channel is vacant but the sensing detected a signal level greater than the threshold. In this case, the

threshold shall be increased by and the new value used as threshold, e.g. if

the channel is occupied, otherwise is vacant.

2) In this situation the database indicates the device is outside the TVWS map and channel is occupied but the sensing detected a signal level less than the threshold. In this case, the

threshold shall be decreased by and the new value used as threshold, e.g. if

the channel is occupied, otherwise is vacant .

Expanding Table 3, the decision process is made as in Table 4.


of 97

Table 4: Final decision process for channel availability based on geo-location and sensing information

G AND S(T) G AND Decision

G = 0 S(T) = 0 Channel is vacant

G = 1 S(T) = 1 Channel is occupied

G = 0 S(T) = 1 G = 0 Channel is occupied

G = 0 S(T) = 1 G = 0 Channel is vacant

G = 1 S(T) = 0 G = 1 Channel is occupied

G = 1 S(T) = 0 G = 1 Channel is vacant

Depending on the choices for and the following cases are possible:

1) AND : Rely on sensing alone with threshold T.

2) AND : Rely on geolocation database only.

3) : Rely on geolocation spectrum database inside coverage area and sensing

outside with threshold of .

4) : Rely on geolocation spectrum database inside coverage area and sensing outside with

threshold of .

These possibilities may be used as policies depending on the region (rely on sensing or geolocation spectrum database) to determine if the channel is occupied or vacant.

Thus, appropriate values for the threshold shall be selected. For example, in the case of wireless microphones, it is required to sense them for values of -126 dBm (and the thermal noise is -121 dBm) or for DVB signals it is required to sense them for values of -120 dBm (and the thermal noise is -105 dBm)

[5]. On the first case, using T = -121 dBm, and , means using a sensing level of -

127 dBm inside the TVWS map area determined by the database and -119 dBm outside the TVWS map area. That approach makes sense because within the TVWS map area devices require extra protection than the ones outside. In a region where both DVB and wireless microphones coexist, the lowest threshold shall be used.

4.2.7- General Architecture Description

In the next description only spectrum commons, which is the one that uses a geolocation spectrum database, is considered. Furthermore, the access to the databases is based on network protocols with reliable connections. Finally, each secondary system is a master controlling several TVWSDs (slaves) with sensing capabilities under the master’s serving area.

When the TVWSD(slave) requires communication, it requests a TVWS channel from the WSD(master);or just requires a channel to start transmitting and no legacy channel is free, so a TVWS channel is required. The TVWSD (master) inquiries the geolocation spectrum database for TVWS availability in the area, providing the GPS location and cell radius1 of the secondary system (WSD master and slave). The geolocation spectrum database transmits the list of vacant TV channels and maximum transmit power (TVWS spectrum pool) to the master.

The TVWSD (master) asks the TVWSDs (slaves) to sense locally these potential channels (only these!) in order to detect wireless microphones, DVB-T or other systems. The WSDs perform the sensing and report the information to the master plus the GPS location. The sensing from more TVWSDs provide a more accurate information; the master combines sensing information coming from different TVWSDs, apply a fusion rule and forward the channel information to be used to TVWSD (slave). The cooperative sensing algorithm can utilize past channel histories to make predictions on future spectrum availability. This process is illustrated in Figure 14.

1 This information depends how the geolocation spectrum database is implemented; the predictions of signal coverage may be

based on transmitted power, antenna type and height, and theoretical transmission models, and not the cell radius itself.


of 97

Under this arrangement TVWS operation is only allowed upon a double affirmative from both the geolocation spectrum database and collaborative sensing. These permits to identify any wireless microphone operations and any other protected signal that might be present at their location but do not appear in the database.

Figure 14: General architecture for combining geolocation and sensing information The diagram presented in Figure 14 is better explained in Table 5.

Table 5: Relevant actions for combining geolocation and sensing information

# Action Required/given information

Description

1 Request for TVWS channels

GPS coordinates of devices, number of channels, occupancy time

The Master detected that needs additional frequencies according to certain location. So it forwards the request to the geolocation spectrum database.

2 Real list of available TVWS channels

GPS co-ordinates, channels, coexistence policies

List of channels per location according to location that are really vacant (from primary users) according to database

3 Request for sensing Device_ID, channel(s) According to the location of each WSD, the Master provides the possible vacant channels and requests sensing on those channels to detect primary users that are not registered


of 97

4 Sensing report Device_ID, channel(s), detected signal strength

The sensing information is sent to the Master to be further processed

5 Compute sensing and geolocation information

Channels, detected signal strength,

threshold T, , .

Detect if channels are occupied or vacant based on geolocation and sensing information. In this phase, a suitable value for threshold T must be selected

6 Notification of channels to use

Channels The Master sends the notification to the Slaves about the channels that shall be used to transmit. There is one channel per Slave

7 Transmission start From that moment on, the transmission may start. Periodical sensing may be required to detect primary users

This proposal for a sensing architecture may evolve in order to follow the work done in other WPs.


of 97

5- Architecture for commons and secondary trading, only geo-location access required

COGEU will consider a centralized topology with a spectrum broker trading with players. The spectrum broker controls the amount of bandwidth and power assigned to each user in order to keep the desired QoS and interference below the regulatory limits. In the COGEU reference model, the centralised Broker is an intermediary between the geo-location database (spectrum information supplier) and players that negotiate spectrum on behalf of spectrum users. The main innovation brought by COGEU is in the combination of unlicensed access to TV white spaces with secondary spectrum trading mechanisms.

Figure 15: Geolocation database with allowing spectrum commons and secondary spectrum trading operations

In line with previous conclusion on spectrum use in section 3-, COGEU proposes to divide spectrum into Unlicensed (on the right side in Figure 15 ) and spectrum market (on the left side in Figure 15). COGEU believe that this is the best way to manage spectrum use, in line with current European regulators situation in early 2011. Indeed, with this assumption as the basis for our scenario, COGEU architecture will benefits from the best of both alternatives (Commons and Spectrum market) while staying flexible enough to fit with any regulator decision in the future.

Figure 16: COGEU reference architecture for commons and secondary trading, only geo-location access required.


of 97

COGEU will consider a centralized topology with a Geolocation Spectrum database dealing directly with TVWS Devices (Spectrum Commons world) or with Spectrum Broker (Secondary Spectrum Market). An overview of the spectrum broker reference architecture is presented in Figure 16. The spectrum broker controls the amount of bandwidth and power assigned to each user in order to keep the desired QoS and interference below the regulatory limits. In the COGEU reference model, the centralised Broker is an intermediary between the geolocation database (spectrum information supplier) and players that negotiate spectrum on behalf of spectrum users. The COGEU reference architecture supports both Spectrum commons and Secondary Spectrum Market. The internal modules of the COGEU broker and its main elements are extensively described in the following sub-sections.

Also the following internal interfaces in the COGEU broker where identified, as shown in Figure 17:

Broker Input Module o Manages the input data from players to the broker

Registration and Validation module o Handles the registration of the players – buyers and validates them when requesting

the broker services

Player communication module o Handles the broker output and communication to the players

The following sections will describe the modules of the COGEU spectrum trading demonstrator.

Figure 17 : COGEU spectrum trading demonstrator (Source COGEU WP7 D7.1)

The Dynamic Radio Engine is a propagation analysis tool based on ray tracing techniques that gets the proposed deployments from players (secondary users) and produces the signal maps for the specific radio environment. It forwards the data to the Coexistence emulator module for interference calculations for the specific area with the specified radio environment. The role of the coexistence emulator module is to quantify the interference between networks operating over TVWS bands. This module will be developed in WP4 (T4.4) and integrated in WP7. As interference metrics the dropped packets and packets retransmissions will be measured under the influence of interference. The emulator is driven by the Dynamic Radio Engine, from which realistic received signal strengths maps for large scale scenarios are supplied.


of 97

The other modules of the COGEU broker are specified in the following sub-sections.

Figure 18 is the Message Sequence Chart of information exchanged between the different systems in order to insure a efficient use of the White-Spaces.

Notify TVWS availability

Demonstration of interest in the TVWS pool

Advertise the spectrum portfolio, call for proposals with initial price

Query (geographic area)

Available and marketableTVWS pool

Query

TVWS portfolio proposal

Interference issues ?

Dynamic

radio engine

TVWS

Allocation

module

Query

Spectrum trading policies

Initial spectrum requirements (area, time and technical operation parameters)

Broker-Player

interface

Bids for TVWS band(s) with proposal prices(s)

Players

TVWS

Device

(Master)

TVWS

Device

(Slave)

COGEU BROKER

Broker-GDB

interface

TVWS occupancy by SU

PlayerRegistration and

validation

BrokerRegistration and

validation

Process TVWS opportunity

Demand estimation

Yes

No

TVWS portfoliocomputation

Benchmark price estimation

Process TVWS opportunity

Auction winner determination

process

Access request

ACK

Access request

ACK

Inform: Start Auction

Enhancement of TVWS pool

Reject Propose Price of some bidders (can be based on trading policies)

Accept Propose Price of some bidders (the winner(s))

Still TVWS to be auctioned ?

YesInform Request Price and End Auction No

AuctionMode

Process the request

Coexistence

emulator

module

Coexistenceanalysis

Generate signal maps for overlaping areas

TVWS

occupancy

repository

TVWS

occupancy

repository

Spectrum policies and

trading repository

Spectrum policies and

trading repository

Geo-location

spectrum

database

Allocate temporarily exclusive rigths

Notification of the channels that shall be used to transmit

New TVWS allocations

Transmission

Player’sRRM

Reconfiguration ReconfigurationCoverage

computation

Update TVWS repository

Decision on channels use

Inform channels to be used

Update trading repository

Figure 18: Message Sequence Chart for COGEU Reference Architecture


of 97

5.1- TVWS allocation and trading mechanism

The spectrum broker determines how the TV white spaces are allocated among players, and also how much each player pays for the acquired spectrum [6][7]. Therefore, TVWS allocation and trading mechanism are important functions of the COGEU broker. Figure 19, shows the cycle of allocation, trading as well as maintenance of the TVWS repository in the COGEU broker. The preparation and analysis phase allocates the TVWS based on a matching algorithm to determine the best combination of the bands and respective technical parameters such as power emission, fragmentation etc, in order to maximize the usage of the TVWS. The trading phase allocates the TVWS to the most valuable users through auction or pricing mechanism. Finally, after the trading phase, the repository is updated to reflect the current status of the TVWS availability. The TVWS allocation and Trading Mechanisms will be presented in the subsequent subsections.

Figure 19: The three phases for the allocation and trading of the TVWS as well as the maintenance of the broker repository

5.1.1- TVWS Allocation

TVWS allocation mechanism: is an optimization problem of resources allocation designed by the technical point of view that can be solved by different optimization strategies as presented in D6.1.


of 97

Figure 20 : TVWS allocation and trading mechanism based on D6.1

5.1.2- Trading mechanism

Several terms are important in the definition of spectrum. These include frequency, geographic location and time. Therefore, for a trade to occur, the needs of the buyer and seller must coincide across all three dimensions [9]. It is not envisioned in the TVWS that the buyer of spectrum will be used to provide DTV services, therefore, there exists incentives for license holders to sell spectrum to other usage as far as there is excess or unused bands which society would obtain more value if it was in the hands of a different user. As a result, the number of participants in a spectrum market may be very low. When the market is ‘thin’, the likelihood that a trade will take place decreases [9][10]. If a trade is conducted, it may occur at a price that is substantially higher (lower) than the buyer (seller) desired. A “thin” spectrum market may prevent current and prospective spectrum users from receiving the price signals they need in order to make decisions on how best to allocate their resources [4]. However, the Broker can increase market thickness by adopting appropriate trading mechanisms that create the opportunity for and enhance the willingness of spectrum users to conduct a trade.

The COGEU broker determines how the TVWS are allocated among service providers, and how much each service provider pays for each spectrum asset. The allocation method, or mechanism, must balance efficiency with complexity [7]. The trading mechanism could be realized through an auction mechanism in which the broker collects bids to buy from the service providers, bids to sell from the geolocation database, and subsequently determines the allocation along with the price for each spectrum asset. The auction would then be repeated as spectrum assets become available (i.e., as they


of 97

are released by supplying players). Figure 21 illustrates the COGEU trading mechanism based on auction.

Alternatively, the COGEU Broker could announce a set of reference prices for the available TVWS, and adjust the prices based on time, location, bandwidth required and other factors to maximize expected revenue or to clear the market periodically. This approach is generally simpler, and requires less overhead (information exchange) than an auction mechanism [7]. However, a well-designed auction mechanism can achieve either a higher efficiency or more revenue depending on the intended objective. As Figure 22 shows, the choice between these two approaches should depend on the thickness of the market. When the market is thin, that is with relatively few buyers and sellers, an auction mechanism may be preferred doe to simplicity in implementation and the chances for higher revenue. However, when the market is thick, with a large number of buyers and sellers, the pricing mechanism should be preferred due to decreased loss in terms of efficiency or revenue. The COGEU broker provides a platform where both pricing and auction mechanisms are supported. In both approaches the algorithm for maximizing profit and the sequences of information exchange for bidding or price adjustments are done automatically. This is supported by electronic definition of spectrum rights.

Figure 21: TVWS trading mechanism based on D6.1


of 97

Figure 22: Thick versus thin market

5.2- Negotiation protocols for trading spectrum rights with the broker

The COGEU broker supports pricing mode and auction mode for allocating spectrum. In the pricing mode, the price is decided by the allocation procedure which considers various factors which influence the value of TVWS in a given place. In the auction mode, the auctioned band has a benchmark price, then each demand (bid) has an associated price and the winning bid decides the final price. In this section, the interfacing signalling between the broker and the spectrum user are presented. The signalling interface is the protocols that enable the transaction of spectrum between the broker and the user to take place efficiently. Through these negotiation protocols, the Broker maximizes its revenue as well as ensures fairness between players. In this case, spectrum is sold in terms of first come first serve basis in the pricing approach, or the most valuable bidder wins the band depending on the auction mechanism. In the following subsections the negotiation sequence steps of the pricing and auction modes will be given. The modes are based on the flow charts presented in Figure 20 and Figure 21.

5.2.1- Pricing mode protocol

Figure 23 gives the operation sequence of the pricing mode protocol. The sequence is as follows:

(1) The Broker informs the players about the available TVWS portfolio and corresponding prices,

(2) Network operators and service providers buys spectrum in first-come-first-serve basis

(3) The Broker authenticates the players

(4) The Broker process the bill for the temporary spectrum rights for the buyer.

(5) The payment systems authorizes the transaction (or payment)

(6) The Broker allocates temporary spectrum rights to the buyer

(7) The Broker updates the local TVWS repository and monitors market


of 97

Figure 23: Pricing mode protocol sequence diagram

5.2.2- Auction mode protocol

The auction mode behaves differently depending on the type of auction. The following is a generic sequence of events in allocation the TVWS through the auction mode:

1. The broker informs the players about the available spectrum.

2. Network operators and service providers send their bids for the spectrum.

3. The broker authenticates the players

4. The broker solves an auction to maximize its revenue or spectrum efficiency.

5. The Broker informs the bid results.

6. Depending on the auction mechanism, an iteration (1-5) continues until the bit winner is found.

7. The broker announces the final results

8. The winner acknowledges the results

9. The Broker process the bill for the temporary spectrum rights to the bid winner.

10. The winning bidder authorizes transaction (or payment

11. The Broker allocates the temporary spectrum rights to the winner.

12. The Broker updates the local TVWS repository and monitors market

13. The player transmit their data.


of 97

Figure 24 shows an example of the English auction protocol message sequence.

Figure 24: English Auction Protocol

5.3- Payment System

The Payment System provides the facility that from the spectrum broker side allows to deliver and check out bills (either repeatedly or only once) from the TVWS users to pay them. In the TV white space context information provision service, the payment system provider plays an important role since the TV white space user is paying every time it wants to acquire reliable TV white space availability information, and hence its functioning should be reliable to avoid being a single point of failure for the model. Assuming a web interface for COGEU Broker TVWS transactions, Figure 25 give an example of payment with an electronic payment system, Paypal. In the figure, the buyer represents the Spectrum Buyer, the Merchants represent the COGEU Broker and Paypal is the payment system. It has to be noticed that secured transaction is an important aspect of the Payment System. In this regards, security is achieved through the SSL protocol as shown in Figure 25.


of 97

Figure 25: Paypal transaction protocol (Paypal) (* Note that before send payment information, Paypal will adjust the buyer and merchant’s Paypal account ) (source [11] )

5.4- External geo-location spectrum database

In COGEU model, the regulatory bodies assign TVWS for spectrum commons (free access) in given areas. The remaining TVWS can be traded in a secondary spectrum market. The design of COGEU geo-location database has to deal with these two operation models. The COGEU geo-location database receives enquires from both, unlicensed WSD’s and from entities running spectrum brokers. [D4.1] Each pixel has associated an operation “Mode” field, in order to make clear if a specific entry will be used by the Broker or by the spectrum of commons model. [D4.1] The main purpose of the COGEU geo-location database is to enable the protection of the incumbent systems from harmful interference. Interference can be caused by white space devices when operated in the same or adjacent channels with the incumbent systems. In order to be able to offer sufficient protection, various parameters need to be specified that will enable and help the protection process in conjunction with the overall anti-interference database design. The design of COGEU geo-location database has to deal with these two operation models. The COGEU geo-location database receives enquires from both, unlicensed WSD’s and from entities running spectrum brokers. In particular, when the available TVWS are calculated, according to the regulator’s policies a percentage of these available TVWS are marked for spectrum commons access, and the remaining for spectrum market. When this distinction is made, COGEU geo-location database model can operate both spectrum sharing regimes. The spectrum commons model will operate in the bands marked for unlicensed use TVWS, and the (COGEU) broker will trade the spectrum that is marked for secondary trading. An entity called spectrum commons manager will make sure that the enquiries of WSD are served from the already divided spectrum for the specific use, and an entity called Broker manager will deal with enquiries coming from the Brokers. The difference is that the Broker will request and receive batch data concerning the availability for all spectrums that is available for trading, and then the broker will use the available spectrum information efficiently through trading.


of 97

Besides the information on the incumbent systems that a database will hold, it will also include the geo-location information per geographic pixel for a specific region and records of the WSD that operate in the specific region. This will lead to database information explosion. Due to the vast amount of information that the database is expected to store, a hybrid approach for the database topology design is required. Therefore, for efficiency and better performance, COGEU geo-location database will adapt a two level database topology. The first level will contain the regulator's controlled information, which includes the incumbent system's parameters. This information can be contained in one database that holds information per country. The second level of information will hold the calculated geo-location information and the operating WSD devices per specified region of control. This design also offers the flexibility of the deployment of more than one database for one region and thus allows competitive operation of database administrators. Details on the interfaces for the geolocation database are given in section 4- .


of 97

5.5- COGEU broker internal repositories

The COGEU spectrum broker role is to determine how spectrum is allocated among secondary users, and also how much each player pays for the acquired spectrum. To do so, it has to maintain internal repositories dealing with secondary users co-existence, spectrum policies implementation, and trading information. In this section, the following internal repositories will be described: TVWS occupancy repository; trading information repository; and policies repository. Details for respective repositories are given in the following subsections.

5.5.1- TVWS occupancy repository

In general TVWS occupancy repository holds TVWS channels occupied by other TVWS networks before the actual allocation process, current status of secondary networks and real-time spectrum occupancy of TVWS. As Figure 26 shows, the TVWS occupancy repository is designed to keep a track of the allocated TVWS spectrum for trading by the broker. In D4.1 it has been described that the available spectrum that the geo-location database management will be divided in two categories. The first category is for unlicensed use, in spectrum of commons, and the second category for trading on the secondary market, which is managed by the broker entity. The contents and operations of TVWS occupancy repository are designed to enable and help the provision of QoS to the secondary players in the TVWS bands. The TVWS occupancy repository is the unit that contains all the information where TVWS devices may transmit and also contains information on active TVWS devices and their operational parameters. The repository carries all the information required to estimate mutual interference between TVWS systems. One of the fundamental parameter in this database is the spatial resolution (the cell size). It is defined once at the introduction of the system and then kept fixed. However, a change to another resolution may not be excluded, even use of different resolutions may be reasonable. The first step is to make sure that correct information is included in the repository. In order for the TVWS occupancy repository to provide updated information to the broker modules for trading, the interface that was described in D4.1 between the geo-location database and the broker is used. The aforementioned interface is designed for exchanging information between the geo-location database and the TVWS occupancy repository. The TVWS occupancy repository is the unit that contains information on active TVWS networks and their operational parameters. The repository carries all the information required to compute mutual interference between TVWS systems (performed by the Dynamic Radio Engine and The Coexistence emulator). The TVWS occupancy repository not only contains the data, it as well hosts the methods to manage the database and generate events or reacts on external events relevant for the management of TVWS systems:

Data

TVWS (channels unused by primary services and maximum allowed transmit power), supplied by the external geo-location database. All TVWS secondary systems in use in the considered area with its describing parameters (Base station position, antenna height, Tx power, antenna radiation pattern and RAT)

Methods

- How to fill the database / updating the database through the Dynamic Radio Engine

- Management of database (add/modify/remove TVWS service)

- Perform interference calculations / calculate coverage of TVWS devices…{using the

methods of the spectrum policy database }

- Handle prioritized services

Events

- Trigger TVWS Update (periodically / on external trigger)

- Start prioritized service


of 97

Figure 26: TVWS Occupancy Repository Communication between the TVWS occupancy repository and geo-location database, are necessary due to the variability of the operation areas and in order to accomplish the validity and uniformity of data between the two databases for optimal protection of the primary users and optimal allocation of bands to the secondary users. There are two different methods for this communication. First, on a specified interval, for instance once a day, the TVWS occupancy repository downloads the available channels for trading. Second, as an alternative, the geo-location database can push this information when and if there is a change in the geo-location database data that affect the current broker trading area. When the information are available at the TVWS occupancy repository, stores them for trading reference, compares the information with the previous and if there are any conflicts regarding the allocated bands, actions are initiated in order to resolve the conflict, based on the predefined policies . The communicated information includes the available channels and the maximum transmit power per geographical pixel that has been marked for trading by the geo-location database entity. When the information for the band availability is in the repository, the broker can use the information for evaluating the secondary player’s requests for usage. During this process when a request is made for usage from a secondary player, an evaluation process is taking place that the current request will not cause interference to the already allocated spectrum. This is important, since the secondary players, request a band and if it is allocated to them, it needs to be guaranteed, under conditions, for optimum availability and interference from other secondary systems. Due to the fact that the broker is solely dealing with the secondary spectrum market, it will only take into consideration the secondary systems that operate in the TVWS bands that the broker controls. As aforementioned, TVWS repository needs to hold information that will ensure the seamless operation of the entity and the broker in general. Therefore, in order to enable the broker to manage the spectrum, TVWS repository needs to include the following information: The available channels with the maximum transmit powers will be in the repository after the processing and inclusion of the current operated secondary systems. The characteristics of the already deployed secondary systems that have been allocated spectrum will also be stored in the repository. Connecting information and expiration dates for the band in use also will be in the repository. These information entities will be needed in order to evaluate current and feature requests from secondary users regarding an available band for usage by a secondary player. Figure 27 shows the structure of the repository and the main entities are highlighted. As can be seen, the main entities are the “TVWS” which holds the available TVWS and the SYSTEMS and PLAYERS which hold the registered respective systems and players. Table ACTIVE_SYSTEMS has the current active systems that are operating under the Broker and have been allocated band through the TVWS allocation process.


of 97

Figure 27 : Broker TVWS occupancy repository database structure. When a request is made by a secondary player, an evaluation process is taking place that the current request will not cause harmful interference to the already allocated secondary systems. This is important, since in COGEU model the Broker has to guarantee temporally exclusive rights for TVWS usage (clean spectrum). The coexistence evaluation is performed by the Dynamic radio Engine and the Coexistence emulator module described in the following sections. Moreover, when a specific band is allocated to a secondary player for usage with a specific technology, the band is marked as “in use” status with the expiration of the licence of usage noted as well in order to avoid the reallocation of the same spectrum during the periods that is already allocated.

5.5.2- Trading Information Repository

In order to facilitate trading of TVWS, participants (Broker and Players) require legal certainty over the rights and obligations that will be transferred. A spectrum usage right is related with the availability of the spectrum (in terms of bands or channels) and ensures that all entities that want to compete for the spectrum usage are able to do it under the same conditions. Obligations specify conditions that players must fulfil in order to maintain their rights. For example, when a specific channel in the COGEU secondary spectrum market is attributed to a certain entity, that entity has the right of exclusive use, but with obligations to protect primary users and limit the interference with other secondary users. In order to enforce these policies it is necessary that records are maintained by the COGEU Broker. As such, the COGEU Broker internal database will keep records of market information which will enable it to monitor and enforce these policies, such as: Spectrum ownership records:


of 97

The past, and present, ownership all spectrum lots will be recorded. For every Spectrum Usage Right (SUR) lot, i.e. spectrum defined by bandwidth, time, transmission power, a record will be maintained. This will enable the regulator to track the historical use of frequencies in particular areas.

Unique spectrum trader or Player records:

For every unique Player, the current (and future) spectrum holdings, i.e. all licences currently held by a unique market entity, and those bought for future use, will be recorded.

Also, past spectrum holdings, i.e. all licences previously held by a unique market entity will be recorded.

The technical and service use to which these licences are put will also be recorded, i.e. Wi-Fi, WiMAX, Public Safety, etc.

·The actual use of licences will also be recorded, i.e. if a Player buys a licence but does not deploy a network to use the frequencies this will be recorded.

Holding this data will enable the Broker to analyse trends in spectrum use and spectrum ownership. It will be able to adjust policies regarding anti-monopolisation behaviour such as spectrum caps and how to handle prioritized services such as public safety applications. It is assumed that all companies will have one unique identity. Analysis of past usage on the part of any individual Player will enable the Broker to profile the Player. Analysis of spectrum usage trends will also enable regulators to amend policies regarding usage rights - emerging or declining technologies and services may demand adjustments to the way in which SURs are defined in licences. Without recourse to historical records it is difficult to analyse trends and anticipate future policy requirements.

5.5.3- Policies Repository

A Policies Repository (see Figure 28) secure the policy management and distribution mechanisms in order to prevent malicious users from altering loaded policies as well as from inserting additional policies and thus causing harmful interference.

GUIRegulatorRegulator

Trading mechanism and Price discovery

Dynamic TVWS allocation mechanism

PoliciesRepository

rulesrules

Figure 28: Policies Repository functional block

Authenticated and authorized stakeholders create and modify policy documents using the Policy Management Tool. This tool (see Figure 29) is used to input active policies and it takes high-level policy information to construct a more detailed, low-level policy description that can be applied to various devices in the network. This should be implemented as an intuitive graphical user interface (GUI) and can be either a standalone application or a component of a mission / network planning system. The resulting detailed policy description is stored in the policy repository (see Figure 28).


of 97

Figure 29: Policy management tool

The policy repository is most commonly a database, files or directory service. When persisted in a directory service, policy distribution becomes a part of the overall directory replication mechanism. This enables the policy descriptions to be automatically distributed throughout the network and aids in the synchronization of policies between the creation points and the enforcement and decision points. Policy enforcement points (PEP) enforce and execute different policies, and can use intermediary policy decision points (PDP) instead of communicating directly with the repository.

Policies are used to describe the constraints on using the spectrum, such as for how long the spectrum can be used, what level of interference is allowed and what power level is allowed to transmit. The policies may vary depending on geographic region and time; therefore the systems must be able to load new policies at runtime.

5.5.3.1 Regulatory Policies for the Specification of Spectrum Usage Rights and Obligations

Rights to transmit or receive signals over spectrum can be defined in relation to five generic parameters:

Geographical area e.g. an entire country, a region or a defined area around a base station. Duration and time of access, e.g., unlimited or defined length of licences, gain access to

spectrum throughout the entire day, or at a specific time of day only.

Spectrum endowment– the frequency bandwidth to which access is granted.

Protection from interference– the right to receive signals without harmful interference from other spectrum users, and the obligation not to cause such interference.

Protection from market abuse position– task here is to strike a balance between allowing abuses of market power and penalising firms which, through their own efforts and innovation, have established a strong market position, but are still subject to competitive threats

All existing spectrum rights, whether distributed under a command-and-control approach or the market model, can be defined in terms of these generic parameters. In relation to the basic spectrum endowment, geographical area and duration/time, it should be quite straightforward to extend these concepts to tradable rights. The spectrum endowment can be specified as a start frequency and end frequency. Geographical boundaries are generally specified as vertical planes between two grid references, where TVWS users can change spectrum use interference needs to be regulated (e.g. one which limits emissions at the boundary) to protect other users from higher priority on a national basis. Duration and time of access are relevant where the right to access spectrum is being transferred between users for defined periods of time (e.g. as it is the case of special events broadcasting) and/or shared between users. Existing licences, when converted to liberalised spectrum usage rights, should retain a definition of the current deployment. If required, such restrictions can be relaxed or removed through negotiation with neighbouring spectrum users. New licences can start with or without a definition of deployment and the


of 97

rights may depend on the approach that is chosen. While systems using highly directional antennas could be accommodated within a general technology-neutral spectrum framework, it is considered that this would likely result in gross inefficiencies. Centralised planning is required for this and it may be appropriate to define a different style of licence. Regarding to this, a clear recommendation made in [D2.2] is that new property rights should be established for the spectrum trading, or more precisely the trading of spectrum licences. The rights and obligations of tradable licence must be sufficiently clear: duration, area and interference restrictions. Obligations for the owner of the spectrum or broker have the following points:

Coverage requirements;

Quality of service requirements;

Interoperability requirements (e.g. handover, roaming);

Minimum service offering (e.g. location-based services, high speed data transfer, video telephony);

Social aspects and universal service obligations, for instance special services for the disabled.

The licensee's obligation not to interfere with other spectrum user’s rights.

The licensee's degree of protection from other users;

The band which is available for use;

The geographical area in which it can be used;

The period for which the licence is entitled. Procedures for scrutiny and reaction by the broker responses must be in place to prevent or avert the consequences of trades which confer high levels of market power on firms acquiring licenses. Where existing licences become tradable and subject to change of use, rights should be established consistent with current uses; this will avoid conflicts of rights and permit parties to renegotiate rights when circumstances change.

5.5.3.2 Rules for Dynamic TVWS Allocation

To improve the utilisation of the spectrum, more dynamic access schemes are now being considered by regulators. This consideration is supported with advances in the development of agile radio technologies. The central idea of dynamic spectrum access is not just that it leads to a great innovation in radio technology, such as cognitive radio, but that such innovation is linked to innovation in the use of the spectrum to meet commercial and social goals. As such, it should not be the aim of regulators to set the rules that are directly enforced in cognitive radio implementation. Instead, regulators shall provide a framework of rules that later are translated to concrete implementations according to cognitive radio systems.

In order to maximise the opportunity of innovation in the combination of software radio technology and dynamic spectrum access policies, a highly disaggregated approach to the delegation of policy-making authority should be encouraged. For example, policy-making authority could be delegated to municipal authorities to best meet local social conditions or different spectrum trading commodity markets could be established in different bands or regions to allow parallel experimentation with market rules. In this case the regulatory supervision is required.

5.5.3.3 Prioritization of TVWS Access

The hierarchy of polices is very important to define prioritization of policies for TVWS. Figure 30, below, represents a possible policy hierarchy and accessing spectrum. Once the TVWS device learns the static policies that apply in any location, it can dynamically resolve user’s requests for spectrum based on more situational policies, dependent on factors such as the application, the user’s role in the incident or the developing incident command system as an incident grows and wanes [12].

Some spectrum access policies may be static, some may be universal, and some may be dynamic or regional. Some policies may only be invoked in certain circumstances, and at certain locations. Some static policies may be hard-coded into the TVWSDs when they are manufactured, while others may be downloaded periodically from a database. We envision a hierarchy of spectrum pool policies, which will guide the user to the best choice for the channel based on its ability to resolve available options within a structure of rules.


of 97

The type of used antennas, like directional, creates space for other links, so could acquire spectrum at cheaper prices. During disaster, all systems should be in emergency mode, so no charges at that time – hence the adjustment factor adds some degree of manipulation to the spectrum band – which could increase spectrum efficiency by attracting or repelling services in certain bands or regions.

Figure 30: Hierarchy of polices

5.5.3.4 Protection of Competition

Regulatory bodies are designed to make decisions in a slow and methodical process that is transparent to everyone, and to promote compromise among competing interests. Such processes are great at avoiding corruption and favouritism, but not well suited to reacting quickly to exploit a new technology, or to repair newly visible interference problem or security vulnerability. Thus, where a regulator is in charge, all else being equal, there are more reasons to select a secondary spectrum sharing model based on coexistence. Spectrum trading allows the assignment of spectrum to respond to changing circumstances and so use spectrum more efficiently [25]. Whilst the provider of a spectrum-derived service might be able to make up for a shortage of spectrum by additional expenditure on infrastructure, however a minimum amount of spectrum is still required. Therefore, if a rival purchases the entirety of this provider’s spectrum, it might be able to eliminate its competitor, reducing competition in the downstream market for services and rising prices. Whether such exclusionary strategy succeeds and has an appreciable effect on competition will depend on the entry barriers to the downstream market. Providing competition is effective; there should be close accordance between the willingness of spectrum users to pay for spectrum and the benefits they generate for end users. One provider will be prepared to pay more for spectrum than another either because it can provide a superior service or because it can reduce other costs. However, if competition is not effective, then it does not automatically follow the statement that the entity prepared to pay more for spectrum is the one that can generate greater benefits for consumers. Some of the willingness to pay for spectrum may derive either from existing market power or from the anticipation of gaining market power. Competition is guaranteed through quantitative limits on spectrum holdings applied at the time of primary assignment. Given that, the secondary trading or liberalisation restricts the limits applied at the time of a possible second assignment, and this limitation shall be forward over the lifetime of the licence, unless further spectrum becomes available. In order to protect competition the following rules must apply:

The regulatory authority establishes rules that specify how spectrum trading should take place.

Trades or transfers of spectrum are subject to approval by the regulatory authority.

Spectrum owned by one company cannot be used by another.


of 97

The spectrum charge should be calculated fairly, i.e. if two users are using the same amount of spectrum in the same way, both should pay the same charge.

The pricing structure should be clear, transparent and comprehensive, without unnecessarily lengthening the licensing process.

In conclusion [24], “concerns that spectrum trading could lead to hoarding of spectrum are not well grounded. In fact, there are a number of ways in which spectrum trading may actually reduce incentives to hoard spectrum. First, liberalisation can relieve spectrum scarcity and may increase competition in downstream service markets. This reduces the need for associated obligations (such as roll-out conditions) to ensure services are delivered to customers. Second, spectrum trading (even without liberalisation) leads to greater transparency of the opportunity costs of leaving spectrum unused. Even if spectrum is unused, such behaviour is not necessarily anti-competitive per se and clearly needs case by case analysis. Such problems appear to be ideal candidates for treatment through standard competition law.”

5.5.3.5 Possible Implementation

The operation of the market according to the open-cry auction model establishes that initially the spectrum seller need to define a policy statement that specifies which banks they trust to provide micropayment tokens and have an account with. Figure 31 [14] presents an example that defines a contract between buyers and banks (public keys and signatures are truncated for readability). A spectrum buyer needs to have a contract with a bank that allows them to spend micropayment tokens.

Figure 31: Definition of the contract between buyers and banks

The spectrum seller specifies in a signed policy statement an offer which is sent to the clearing house. The statement specifies the terms of the offer and the signature binds the seller to these terms. Figure 32 [14] is a keynote credentials the details of the offered spectrum usage rights, the expiration date of the offer, and that a bid should be greater or equal to 3.0 monetary units.

Figure 32: Definition of the parameters to seller


of 97

When a seller finds a bid they first need to verify that the buyer can indeed deliver the payment and that all market / context / technical terms are satisfied by the given bid. If the buyer has a valid contract with a bank that the seller trusts and the offer terms are met by the proposed bid then their PDP authorizes the transaction (Figure 33 [14]). Therefore, the seller is sure that they are going to be paid for the provided spectrum usage rights.

Figure 33: Definition of the parameters to buyer

When the seller informs the buyer that they accept the bid the latter releases the required number, in the above case 30, of tokens to the former. At that point, the seller issues a signed keynote policy statement to the buyer that includes the purchased spectrum usage rights as specified in the initial offer. It is called the spectrum use credential. It can be used by the buyer as a proof of the spectrum assignment facilitating the operation of policy schemes. The seller collects tokens and periodically contacts the issuing bank to translate them into monetary units and deposit them to their account.

5.5.3.6 Conclusion The exclusive right to operate a transmitter within a given frequency range in a given geographic area is only part of the definition of a flexible spectrum license and once the authorization process is underway the role of the regulator is to ensure non-discriminatory treatment of all players in the liberalized market. For COGEU the right and obligations shall be clear and transparent, and additional obligations on the licensee must be specified to limit interference. In addition to the rights of transmission, the repository also stores the rights and obligations of user of communications services. The user has a right to have access to communications services (e.g. voice, data, etc) at reasonable prices. The user have a right to receive a service with a quality that reflects the cost of the service (value for money) and also the user have a right to receive the level of quality of service that is quoted or stated by the service provider/operator in the Service Level Agreement(SLA) [D2.2]. In conclusion, the user has a right to complain about quality, delay, quantity and tariff with regard to the nature of the communication service provided.

The temporary owner of spectrum may issue Spectrum Rights (SURs) as they see fit so long as the conditions of the Spectrum Management Rights (SMRs) are met. Otherwise it would be necessary for the SMRs owner to negotiate a change of its SMRs parameters with any affected neighbours. It is proposed that use of spectrum by an SMR holder should be registered through issuing itself with SURs, so as to reduce transaction costs and improve spectrum efficiency.

Buyers and seller require clarity over the expiry of usage right. If the duration of a usage right is uncertain, or approaching its end date, this will depress the value of the licence in a secondary market. Ofcom distinguish between the four following types of transfer for spectrum use [16]:

Outright total transfer: An outright transfer of all the right and obligations arising under a licence to a third party.


of 97

Concurrent total transfer: A transfer of all the rights and obligations to a third party which results in a concurrent holding of those rights and obligations by the transferor and the transferee(s).

Outright partial transfer: An outright transfer of some of the right and obligations arising under a license to a third party.

Concurrent partial transfer: A transfer of some of the rights and obligations to a third party which results in a concurrent holding of those partial rights and obligations by the transferor and the transferee(s).

The traditional framework is highly prescriptive and often requires very detailed regulatory oversight. For example, it may prescribe the applications that can use spectrum (e.g., mobile services, terrestrial point-to-point links, etc.), the technology to be used, power levels, localization and height of the transmission masts, as well as bandwidth. Liberalization means removing, or at least reducing, these restrictions. Introducing liberalization and spectrum trading in parallel would facilitate the migration of spectrum usage rights to more efficient uses. This boosts efficiency, furthers innovation and makes competition more intense. However, in so doing, it must be considered that specific restrictions are necessary to avoid harmful interference, while other requirements are necessary to satisfy international agreements.

If a country decides to introduce spectrum trading, it must:

Establish clear legal rules to spectrum rights so that buyers and sellers have certainty about what it being traded, and these should enable leasing (both short term and long term) as well as the complete transfer of rights;

Publish information about what spectrum has been allocated in a public register, and what transfers have taken place

Establish a quick and simple process that enables the regulator to prevent undesirable trades.

Beyond the rights and obligations, the COGEU policies repository also includes different priorities. The prioritization mechanisms for public safety are set to be the highest priority in case of disaster. Under such conditions; all systems should be in emergency mode, to guarantee reliable communications between the public safety workers. At this point it is clear the necessity to allow national regulators to control the TVWS spectrum assignments. However, it shall be taken into account the fact that polices of prioritization for public safety is different from country to country.

5.6- Graphical User Interface

The purpose of the Graphical User Interfaces is for the communication and interaction with the secondary players. In order for a secondary player to communicate with the broker, a graphical user interface is needed that will ensure the information passing between them are verified and valid. There are three main steps in the interaction process. The first step is for the broker to advertise and make public to the potential players that a spectrum or a specific band is available. This can be done by using RSS feeds technology or by email technology. This can be accommodated in a user graphical interface by providing an RSS feed or an internal email client. The proposed method has the characteristic that only the registered players can view some of the advertisings and some can made public through a web interface that a public access is in place. The second step is for a player to put its interest forward. In order for a player to clearly define what and where wants a network to be deployed, a map of functionality and system characteristics input will also be available in the graphical user interface. The system will need to input with very high accuracy in order to be able to evaluate the request for usage in a specific location. The third step is the answer from the broker system. When a player indicated where and what exactly wants to deploy in the available band, a validation and allocation process will be held in the broker and the player will be able to observe and inform regarding the outcome of its request using the graphical user interface.


of 97

For a successful usage and communication between the broker and the players the graphical user interface need to stay simple and intuitive and also hold help functionality and troubleshooting capabilities.

-

5.7- Interfaces for geo-location database access and regulatory enforcement

The interfaces of the Geo-location database for this regulatory assumption are the same as for the Spectrum Commons regulatory scenario (presented in Figure 12 in chapter 4-) with the addition of a new Access Domain for the COGEU Broker. Figure 34 presents the four access domains of the geo-location database, for every domain different access interface is available with different access rights and limitations. This has been designed with the security and data integrity in mind. Commons Public and Access domain can only read the database for different purpose and Regulatory access domain can update and alter the database. The major addition compared to the 1

st regulatory scenario is the

Broker Access Domain.

Figure 34 : Geo-location database access domains.

5.7.1- Interfaces and Protocols for Geo-location database access

Database enquires and downloads use appropriate protocols. COGEU will adopt Internet-based protocols and standard enquiry languages. The proposed database access procedure includes XML through web services. In more detail the database will expose web services though the appropriate interfaces. The transported data will include XML formatted data and by using SOAP encapsulation will be transported through HTTP. Using the above methodology, the architecture has the benefit that the access to the database is controlled, and secure, since the queries that an entity will be able to perform are predefined and preformatted. Through this process the entity that accesses the database can be authenticated and authorised according to the policies that are in place at the current time in the specific database. Also the benefit of well formatted data that can be easier to manipulate and validated is inherited through the use of XML. The database is protected to the boundaries of the exposed interfaces, simply put no entity can directly access the database and alter stored information.


of 97

An entity will be able to use the services of the database, by connecting to the appropriate exposed port of the database and using the available queries. On the other hand the respond from the database will bear the same characteristics, which will provide well formatted and easier to manipulate data to the entities. That will add the benefit of faster processing of the transported data between the databases and the querying entities. In general faster processing of the exchanged information will be possible using the proposed process.

Figure 35: Example of list of available channels format reported by the geo-location database for a specific location

5.7.2- Interface for populating the geo-location database

In order to populate the geo-location information in the geo-location database, the operation need to use the Interfaces C, D and F. Interface C is for retrieving regulatory information, Interface D is to retrieve the incumbent system characteristics and Interface F is to compliment the process if needed with information held in the central database. COGEU assume a realistic scenario where the regulators will not supply the sensitive data concerning broadcast transmitter parameters. Therefore, the regulator would convert the incumbent’s data (confidential raw data) into a list of allowed frequencies and associated transmit powers by performing TVWS calculations. As a result of these calculations, regulators may use a map with a grid size of e.g. 200 m x 200 m (‘pixel’). For each pixel and each channel the acceptable transmit power is contained in the database as shown in Figure 36. In COGEU, these TVWS calculation for the Munich area will be performed based on the data given by Germany’s broadcasters and German regulator BNetzA. For the considered Munich area and a pixel size of 200 m, for each channel arrays of approximately 2000 x 2000 items will be supplied e.g. as ACSII files (in this case the whole federal state of Bavaria is covered). COGEU database will be populated by inputting these data. COGEU assumes that a database for PMSE is either available or will be built up in advance of introduction of white space using equipment. However, PMSE use in Germany is not registered. To get a realistic scenario for COGEU demonstration in Munich some assumptions have to be made on distribution of PMSE equipment2. For example, it can be assumed that broadcast production companies, Theatres, Stadiums and Universities use such PMSE systems. By estimating the number of devices used at these facilities some channels may be excluded for these locations.

2 COGEU has demonstrated a PMSE booking tool in the COGEU Workshop which took place in

Munich, on the 10th of November, 2011.


of 97

Figure 36: The process of TVWS computation.


of 97

5.8- Interface between the COGEU broker and players

Standardisation efforts are ongoing to specify air-interfaces for use in white space spectrum. As introduced in Section 2-, IEEEs 802.11af task group is currently working on such a specification while IEEE 802.22 is another example already standardised. As COGEU reference scenarios, LTE, wifi and public safety over White spaces are also considered as future expected air interfaces. A strong requirement taken into account in the definition of the COGEU architecture is that the use of white spaces available spectrum is enabled via the capability of a device to query a database / broker and obtain information about the availability of spectrum for use at a given location. The broker is considered reachable via the internet and the devices/entity querying the broker are expected to have some form of internet connectivity wirelessly using frequencies outside the band 470-790 MHz, or via some wired connection in case of non-mobile TVWSDs on a dedicated communication channel. Indeed, we made the choice not to rely on a Cognitive Pilot Channel to support the user terminal for discovery of available radio access and reconfiguration (as introduced in ETSI TC RRS WG3). The advantages of using the geolocation spectrum allocation via the broker and communicating with it are mainly its simplicity of implementation compared to a full-CPC oriented approach. On this communication channel, COGEU adopts a “Service Oriented Architecture” (SOA) communication pattern in order to provide flexibility and standardised interfaces. The main technologies to be used are based on the Web-Services standards:

XML (Extensible Mark-up Language) for data encapsulation

SOAP (Simple Object Access Protocol) to transfer the data

WSDL (Web Services Description Language) described how to interface with a given web service

Figure 37 : COGEU Web-service A messaging interface between the white space devices and the broker is required for operating a network using the white space spectrum. The following sections discuss various aspects of such an interface and the need for a standard. COGEU messaging interface is considering the IETF PAWS (Protocols to Access White Space database) Working Group recommendations. The IETF PAWS Working Group expects to:

Standardize a protocol for querying the database, which includes a location sensitive database discovery mechanism and security for the protocol, and application services.

Standardize the data structure to be carried by the query protocol. Since the location of a user device is involved, privacy implications arise, and the protocol will have to have robust security


of 97

mechanisms. Existing IETF location data structures and privacy mechanisms may be considered for use.

It states that an efficient messaging interface needs to be:

1. Radio/air interface agnostic - The radio/air interface technology used by the white space device in available spectrum can be 802.11af, 802.16, 802.22, LTE etc. However the messaging interface between the white space device and the database should be agnostic to the air interface while being cognizant of the characteristics of various air-interface technologies and the need to include relevant attributes in the query to the database.

2. Spectrum agnostic - the spectrum used by primary and secondary users varies by country. Some spectrum has an explicit notion of a "channel" a defined swath of spectrum within a band that has some assigned identifier. Other spectrum bands may be subject to white space sharing, but only have actual frequency low/high parameters to define protected entity use. The protocol should be able to be used in any spectrum band where white space sharing is permitted.

3. Globally applicable - A common messaging interface between TVWSD and databases will enable the use of such spectrum for various purposes on a global basis. Devices can operate in any country where such spectrum is available and a common interface ensures uniformity in implementations and deployment. Since the White Space device must know it's geospatial location to do a query, it is possible to determine which database, and which rules, are applicable, even though they are country specific.

4. Address regulatory requirements - Each country will likely have regulations that are unique to that country. The messaging interface needs to be flexible to accommodate the specific needs of a regulatory body in the country where the white space device is operating and connecting to the relevant database.

In a typical implementation of geolocation database to access TV white space, a radio is configured with, or has the capability to determine its location in latitude and longitude. At power-on, before the device can transmit or use any of the spectrum set aside for secondary use, the device must identify the relevant database to query, contact the database, provide its geolocation and receive in return a list of unoccupied or "white space" spectrum (for example, in a TV White space implementation, the list of available channels at that location). The device can then select one of the channels from the list and begin to transmit and receive on the selected channel. The device must query the database subsequently on a periodic basis for a list of unoccupied channels based on certain conditions, e.g. a fixed amount of time has passed or the device has changed location beyond a specified threshold. A proposal data model is currently available as a Internet Draft in the PAWS WG “Protocol to query a White Space Database” [40]. A data model for the query/response protocol is proposed in this document. A hierarchical object model approach is used for defining the query/response and its attributes.

Figure 38 : Data Model Proposal


of 97

The data model includes two main objects, the wsQuery and wsResponse objects. Each of these objects contain a list of elements. The elements are further comprised of attributes. The wsQuery object is sent by a WS Master device to a database and the database responds with a wsResponse object. The actual message and header in which this object is carried is not specified here and is expected to be specified elsewhere. While more information about the nature of each elements and attributes are available in [40] the following illustrate, as an example, the definition of the location element:

Figure 39 : Location Element The complete list of element proposed in [40] for the Query and Response objects includes:

Station element

Schedule element

ChannelList element

ContactList element

Location element

Antenna element

StationRxList element

TransmitterList element

Address element

Coordinate element

RadiationPattern element

Contact Element

Extension element

WhiteSpaceFrequencyList element

WhiteSpaceFrequency element

Channel Element

Transmitter Element More detail about the nature of each elements and attributes are available in [40]

5.9- Registration and validation in the COGEU broker

In the secondary spectrum market implemented by COGEU, a security mechanism is necessary to protect messages transmitted between each player and COGEU broker. These messages are control messages to request or assign TVWS carriers, update information about the market, etc. Security issues between COGEU broker and geo-location database will not be treated here, however a security channel may be assumed. The importance of security may be seen at least in three situations: 1) since the COGEU system assigns TVWS channels and the spectrum is a scarce resource, security attacks may lead to occupancy of the spectrum from those that do not have the right to do it; 2) in the secondary spectrum market, fake entities would request more TVWS carriers and possibly resealing them, building a parallel market; 3) fake entities would notify players has being the COGEU broker, and then collect money from the spectrum and spectrum rights that they do not own.


of 97

As previously mentioned, the information exchange between the COGEU system and players is highly sensible. However, simple and standardised mechanisms are proposed: the asymmetric cryptographic or public-key cryptography. One of the strong benefits of this method is that it allows the transmission of sensible data through Internet in a secure way without establishing a security channel. It is assumed that each player and COGEU broker has a public/private key pair. It is also assumed that only the owner of the private key knows its private key, while the COGEU broker public key can be broadcasted in the network through a security channel (or other security mechanism). When players establish a contract with the COGEU broker allowing them to request TVWS carriers and participate in the market, their names or IDs and corresponding public key are stored in a certificate, which can be signed by the COGEU broker. In other words, the COGEU broker may be the one that generates valid certificates for the players; information about player’s public key, its ID, key usage, that may be delivered by the player to COGEU system when the contract is established, is later included in the certificate. Other important information that is in the certificate is the issuer (the entity that generated the certificate), the validity (valid from and valid to) and serial number (that is used to uniquely identify the certificate). It is not necessary to store the certificate inside the broker, because its validity can be checked. The COGEU broker signs the certificate directly (as explained before), or a certification authority (CA) (more complex with probably no extra benefit) signs the certificate, establishing trust over the public key. For this purpose it seems sufficient that only the broker signs the certificate, since it then can revoke it when needed, and assert more control and trust over what players get signed certificates. If a CA is used, IDs and certificates must be stored inside the broker. Players obtain the certificate of the COGEU broker through some off band/secure mechanism. Alternatively, if received through a public channel, it can verify the signature, provided that the certificate is signed by a trusted authority (CA), or the by the provider (for which the player should have the public key – this can be distributed with the device, or by off line/band mechanisms). In Figure 40 the Authentication and Authorization architecture is presented. The Authentication Server (AS) is in charge of generating the certificates for the players and broker, verify the authenticity of the requests, and sign/encrypt the response. The players that may request TVWS channels from the broker are stored in the Policies Repository, when a CA is used.

Figure 40: Authentication and Authorization Entities in COGEU system


of 97

To actually perform a request, the player first generates a request and a nonce (a large pseudo-randomly generated number), to assure the freshness of the request. The player then generates a hash of the concatenation of nonce with request. The hash is then signed with the player’s private key, so to assure the origin of the request. The player then takes the nonce, request and signed hash, encrypts the package with the broker’s public key, thus forming the encrypted request (Eas). The player then transmits the encrypted request along with his public key, to verify the signature, and with the public key certificate, so that the broker can verify the identity of the sender.

Kop, Cert, Eas (nonce, sreq, SIGN(Hash(nonce, sreq))) ->

The sreq message is constituted by the following parameters:

Bandwidth

Location of secondary devices (e.g. GPS coordinates)

Maximal transmit power

Occupancy duration When the broker receives the request, it simply decrypts the request with his private key, thus obtaining the nonce, request and hash. To verify the validity of the request, the broker first verifies the public key certificate, assuring the senders identity, generates the Hash(nonce, sreq) and compares the obtained hash with the signed hash. If there is a match, the request is valid, and was signed by the true holder of the private key, corresponding to the received public key (verified through the certificate), see Figure 41.

<- Eop(nonce, sresp, SIGN(Hash(nonce, sresp))

To issue a response, the broker does a similar process, but using his private key and the public key of the player. The broker computes the hash of the response plus a fresh nonce, and signs the results with his private key. It then sends the encrypted response to the player, which is the nonce, response and signed hash, encrypted with the player’s public key, thus assuring that only the player is able to open the response (Eop). To validate the response, given that the player already possesses the verified public key of the COGEU broker, it simply decrypts the response with its private key, calculates the hash, and verifies the signature, thus confirming the validity and freshness of the response.


of 97

Figure 41: Example of signing a message

The response (sresp) will be later defined. However it shall contain if the request is denied or accepted and, in the latter case, the number of channel(s), frequencies, policies, duration, and other information that is necessary.

5.10- Mobility management architecture features to support seamless handovers in heterogeneous environments

In COGEU, mobility management has to deal with changing of frequency bands availability (spectrum mobility). Mobility management strategies in COGEU radio network are paramount for seamless operation across different TVWS bands and diverse radio access technologies (e.g. cellular LTE over TVWS, WiFi) over TVWS. This work reports features which are important for mobility management. Further, integration with the Media Independent Handover function in IEEE 802.21 standard is envisaged, extending the traditional mobility concept. The work extends the work presented in D3.1 focusing on the spectrum mobility management aspects under the spectrum commons: both for joint sensing and geo-location and only geo-location access regulatory scenarios.

5.10.1- Introduction

In Chapter 3 of COGEU D3.1, the system requirements for mobility management in heterogeneous networks operating in the TV white spaces were presented. It was seen that mobility management is vital for realizing employable cognitive radio wireless networks to provide diverse cost-effective services such as broadband Internet access, public safety, mobile TV, electronic news gathering, health monitoring, etc. Further, mobility management for cognitive radio wireless networks, including the spectrum aware mobility support remains largely unexplored. Therefore, challenges arising from cognitive radio wireless networks in TV white spaces call for the design of new mobility management techniques considering the environmental features that influence connectivity in temporal, spatial and fractal domains.


of 97

Figure 42: Possible scenario of hand-over across heterogeneous networks in TV white spaces

This section intends to include mobility management features to support seamless handovers in heterogeneous environments in the final COGEU architecture. In this section, mobility management architecture for COGEU use cases in TV white spaces is developed while exploiting context information in the COGEU geo-location databases as shown in Figure 42. Current literature classify mobility into four categories (or levels) [31][32] namely (1) user mobility, (2) terminal mobility, (3) session mobility, and (4) seamless mobility. Details are summarized as follows:

1. Terminal (device) mobility: In this case, the user moves his terminal or device across different networks and access technologies without being disconnected.

2. User (personal) mobility: In this case, the user is able to access services through any terminal, anytime, anywhere. The identity of the user does not change with the change of a device and respective network point of attachment (PoA).

3. Session mobility: In this case, the media of an ongoing communication session is seamlessly transferred from one device to another.

4. Service mobility: In this case, the network is a able to offer personalized services regardless of the end user's point of attachment while maintaining the ongoing session's QoS. Moreover, the network is able to ensure that the user has access to all of its subscribed services and features regardless of its PoA.

Furthermore, with the advent of secondary spectrum usage, and especially for the COGEU use cases in TV white, there is an added type of mobility namely:

5. Spectrum mobility: In this case, a cognitive radio user voluntary or involuntary dynamically changes its operating frequency band [27]. In order to allow the cognitive radio to dynamically operate in a non-interfering frequency band, the cognitive radio networks must provide mechanism to maintain seamless connectivity when the change to a new spectrum is taking place. This could mean accessing a different network though a different terminal interface all together [28][29], in our context, could be different COGEU use case networks. Even though this is none of the above mentioned mobility types, however, triggering a spectrum hand-over event may result in activating any of the above mobility types functions [35]. For example, change of interface due to spectrum mobility may mean change of access technology and hence change in session route and PoA.

In COGEU use cases in TV white spaces, the spectrum mobility aspect affects the whole protocol stack. This directs us to approach this problem from a context awareness point of view, so that the spectrum trigger will be treated as context information in hand-over management. This will later be coupled with the MIH mechanism to allow seamless hand over between heterogeneous networks in TV white spaces.


of 97

The problem of mobility management in the context of secondary spectrum usage has been investigated in various works like [27],[28],[29], etc. However, most of this works did not consider the current advancement in regulatory policy regarding geo-location databases, which reduces the devices complexity on spectrum sensing and increases the reliability and predictability of spectrum resource availability. For this matter, more spectrum efficient mechanisms can be developed to exploit the regulatory change and advance the realization of possible secondary users’ networks such as COGEU use cases. The information offered by the geo-location database will be exploited in mobility management in a context awareness perspective.

5.10.2- System Requirements

The mobility management problem for system operating in TV white spaces has to consider a number of key issues both by regarding theoretical and practical TV white space specific characteristics such as fragmentation. The system requirements for the mobility problem in TV white spaces as presented in D3.1, include but not limited to the following:

5.10.2.1 Operative interworking architecture

This issue refers to regarding the design of a mobility enabled IP networking architecture, which contains the functionality to deal with mobility between diverse accessed technologies. In the context of COGEU use cases, the IP based networking architecture will be exploited in extracting geo-location information on TV white spaces availability. The information will further be utilized in facilitating fast and seamless handovers for COGEU use cases users.

5.10.2.2 Optimal choice of access technology Given that a user may be offered connectivity from more than one technology at any one time, mobility management in COGEU has to consider how the terminal and the network choose the technology suitable for services the user is accessing. Information from the MIIS sever will be used to choose the best available Candidate Master Device for handover.

5.10.2.3 Media-independent handover (MIH) IEEE 802.21 is an emerging IEEE standard that has begun in 2004 and is still under development to afford interoperability between several types of wireless access networks. This standard focuses on handover capability between jointed wireless data access technologies regardless of the type of access medium called media-independent handover. Although the IEEE 802.21 standard is still in its formative stages, many frameworks emerged based on this standard, one of which MIH, will be exploited in mobility management for COGEU use cases.

5.10.2.4 Power saving

The mechanism to save battery life of mobile terminal in COGEU use cases is another aspect that benefits from the proposed mobility management solution. The power consumed by the device in searching for spectrum availability is directly proportional to the time spent in hand-over latency and spectrum sensing, as well as transmitted power. For example, the UMTS/LTE use case transmitted power is much higher as compared to WLANs. In COGEU context, one viable approach for saving battery power is to allow joint sensing and geo-location database, thereby spectrum availability information can be obtained from the geo-location and the mobile device senses only bands identified by the geo-location database and hence save the mobile device from spending a lot of power in acquiring spectrum through sensing more channels than necessary. Hence, combining sensing and geo-location database under the first regulatory scenario will lead to power saving in user devices and the network in general as compared to pure sensing. The same applies for the secondary regulatory scenario where only geo-location access is required under the commons regime.

5.10.2.5 Security consideration As different access networks in the COGEU use cases have their own security protocols and controls, maintaining the overall security and privacy is challenging but essential issue. An efficient mobility solution should provide comprehensive mechanism to protect the flow of data during roaming the host as well as context transfer. However, the security consideration is out of scope of the COGEU project.

5.10.2.6 QoS consideration Different applications require diverse level of QoS. Applications such as VoIP and multimedia stream are sensitive to QoS level. Mobility management in COGEU use cases is another important issue towards improving QoS parameters and ensuring seamless connectivity. Providing robust service in


of 97

terms of bandwidth and vertical handover delay in the use cases still remains an interesting and challenging problem.

5.10.2.7 Context awareness The choice of multiplicity of access of users requires complex QoS support which depends on user applications, available technologies, current network conditions, spectrum availability and user contracts. These conditions are known as context. When context information is considered in mobility management, it will be possible to achieve “make-before-break” hand-overs in spectrum uncertainty conditions and thus maintain seamless connectivity.

5.10.3- Mobility Management Architecture Features

In this Section we use spectrum availability, user's, networks, and regulatory environments as a source for spectrum aware context information that can be exploited to support seamless mobility management for cognitive radio networks. The context information may be maximum power limits set by regulators, spectrum holes availability, user QoS preference, type of service provided by the network operator etc. Any of changes in the context information may trigger an extended hand-over event. In other words, changes in spectrum availability, regulatory, service availability, user mobility and preference may prompt changes in operation frequency, access point, service point, type of service etc. To achieve extended hand-over functionalities, this work proposes spectrum context aware mobility management framework to address functional issues including:

Spectrum context acquisition;

Context matching (resource allocation or assignment); and

Mobility management. Figure 43 shows the functional components for spectrum context aware mobility management. The following subsections give the details of these functional components.

Figure 43: Functional components for spectrum context aware mobility management

5.10.3.1 Context Acquisition The component of spectrum context acquisition is responsible for the detection, collection and storage of context information. This information is basically stored in the 3rd party Mobility Management Entity as shown in Figure 44. The spectrum etiquette context information can be obtained from the local geo-location database. The context information of the environment surrounding the user can be obtained from on-the-device mounted sensors like (spectrum, location, speed, inclination, vibration, etc). Wireless sensor networks in the vicinity of the user may also complement the environmental context. The environmental context can be used to detect complex events, like disaster (e.g. by using temperature, location, inclination, etc) and adapt to this new context for example by activating the safety RAT and connect to early responders network supplying location information to hasten the life saving exercise. Spectrum context is given highest priority for life saving activities. Similarly, context information can activate hand-over from LTE to WiFi depending on user-preference satisfaction context, by using network service context. In this work, context information on the availability of WiFi networks, obtained from the MIIS server, is used by the Slave Device to trigger a handover from serving LTE network to WiFi network based on the user’s preference.


of 97

5.10.3.2 Context Matching

In context matching, context information with user's needs for resources are matched. Accordingly, either real-time or on-demand context update service maybe provided -- which in turn may trigger different service provisioning events like resource reservation in case of anticipated mobility changes; spectrum hand-over on incumbent approaching context; availability of new spectrum resource or services availability etc. Once the context is matched to the requirements, an intermediary entity dynamically connects to the e-commerce server to execute retail or business to business (B2B) electronic spectrum context information transaction.

5.10.3.3 Mobility Management The Mobility Management function of the 3rd party Mobility Management Entity performs the actual management of seamless connectivity based on the information provided by the spectrum context repositories, such as the MIIS. It aims at responding to the context provided by the repository to trigger mobility management function based on identified (and reserved) resources for the target class of users or session. The mobility management component allows seamless mobility hand-over in the spatial, temporal, and fractal domains, independent of technology specific limitations. It executes the handover task by establishing a connection to between slaves and masters in case of master-slave architecture. Furthermore, the handover accounts for QoS provisioning based on the tariffs consented by the user for a given service segment.

5.10.4- Overview of the MIH Framework

In this work we consider that the context aware features represented above (5.10.3-) can be used to realize seamless connectivity in the TV white spaces through the Medium Independent Handover Function of the IEEE 802.21 standard [30][32][33]. The scope of the IEEE 802.21 standard is to facilitate the handover between IEEE 802 and non-IEEE 802 access networks (e.g., cellular networks) in a way that is independent from specific access network technology, i.e., achieve medium-independent handover. This is achieved through a mechanism that allows explicit reporting of remote or local link status which triggers events related to mobility management’s entities for processing.

The standard consists of:

i) a framework that allows seamless transition of Mobile Node (MN) between networks with (possibly) technology;

ii) a new entity called MIH function (MIHF); iii) a MIH Service Access Point (SAP), called MIH_SAP, and its associated primitives; iv) a link-layer SAP, called MIH_LINK_SAP, for each network technology; v) an abstract media dependent interface that provides transport services over the data plane on

the local node, called MIH_NET_SAP; vi) a new entity called MIH user (MIH_USR), which is the functional entity that employs MIH

services.


of 97

Figure 44: MIH Reference model and SAPs The core component of this functionality is the MIH function (MIHF) which is located in both the mobile node (MN) and the network node protocol stack. Figure 44 illustrates the MIH reference model. The MIH_LINK_SAP is used to collect link information and control link behaviour during handover. For each existing network technology, such as COGEU use cases, amendments are needed to interact with the basic primitives defined in the IEEE 802.21 standard. Such work is being carried out for the IEEE 802.3, IEEE 802.11, IEEE 802.16, 3GPP and 3GPP2 as well as RRS, etc, in their respective task groups or technical committees. On the other hand, it is envisaged that new technologies will incorporate suitable primitives for IEEE 802.21 natively.

The MIHF provides three types of services as follows:

The Media-Independent Event Service (MIES);

The Media-Independent Command Service (MICS); and

The Media-Independent Information Service (MIIS). Further details are given as follows.

Media Independent Event Service (MIES): The events are notifications that are generated by the link-layer, typically involving the link quality and status. A MIH user can subscribe to receive such notifications, both from the local lower (as shown in Figure 45) and from remote entities (as shown in Figure 44), through the local MIHF. It is possible for multiple users to register to the same event, in which case the notification is sent to all the subscribers. Media Independent Command Service (MICS): Commands are sent from the MIH users to the lower layers. Commands are used to configure, control, and retrieve information from the lower layers. Like events, also commands can be directed to remote entities through the local MIHF. Commands follow a top-down direction, whereas events follow a bottom-up direction. Commands could be requirement to configure a network device, scan available networks, etc. Media Independent Information Service (MIIS): It defines a framework for acquiring, storing, and retrieving information useful for handover decisions within a geographical area. Support for many Information Elements (IEs) is included to encompass the different types of mobility and supported technologies. Static link parameters, like channel information, or medium access control (MAC) address of the access point (AP) are such information

A link-layer element which can provide the MN with network access is called Point of Attachment (PoA). The PoA currently in use by a MN is said to be its serving PoA, while the PoAs under evaluation for


of 97

possibly becoming PoA in the near future are called candidate PoA. Furthermore, the PoA that has been selected to become the next serving PoA is the target PoA. The MIHF of the MN exchanges MIH messages with a peer in the network that is called its Point of Service (PoS). For maximum flexibility, the communication protocol between MIHF entities is specified by the standard for both layer 2 and layer 3: layer 2 transport is allowed with the EtherType value set to that for MIH Protocol; layer 3 transport is supported for the Transmission Control Protocol (TCP), the User Datagram Protocol (UDP), and the Stream Control Transmission Protocol (SCTP). An acknowledgment service can be enabled to add reliability to message exchange if the transport method adopted does not already provide this.

Figure 45: Architecture for integrating the Media-Independent Handover framework with COGEU user cases and the mobility management component

The IEEE 802.21 standard is limited to handover initiation and preparation phases, while execution phase is not covered, and hence mobility management in upper layers is still required for service continuity across networks. Moreover, the adoption of 802.21 solution in non-IEEE 802 technologies would require the allocation of functional MIHF entities in the terminal and within both wireless networks along with the corresponding transport capabilities to exchange MIH protocol messages. In this sense, MIH signalling to and from terminals has to be realized in the radio link layer, an issue which is not supported by all groups, for example the 3GPP has not yet considered its inclusion in release 8. On the other side, the Internet Engineering Task Force Mobility for IP Performance, Signalling, and Handoff


of 97

Optimization Working Group (IETF MIPSHOP WG) is specifying the higher layer transport for the MIH protocol and mechanisms to discover MIHF peers [30]. This means that MIH is still an attractive technology for enhancing IP mobility across heterogeneous accesses for future internetworking. In this work we focus on exploiting the MIH framework through integrating context awareness framework with the network protocols, as Figure 45 shows, in order to address the mobility management in the TVWS through the geo-location database. The key MIH benefits [34] include:

Coverage extension and preferred networks, as different radio access technologies can be combined to provide seamless connectivity to mobile users with converged devices. The MIH client algorithms enable handovers to a user’s preferred network.

Load balancing and capacity increase, for example, by off-loading a 3G network when hotspots are available. Traffic is spread across different radio networks to balance network loading, while increasing the number of users that can be supported across the entire access network.

Cost efficiency and simplicity of new network roll-outs, leveraging the existing radio infrastructure of cellular 2G/3G, WiFi and WiMAX networks, new hotspots and base stations – using any air interface – can be added dynamically to scale the overall radio network footprint and capacity of these combined technologies.

Flexible network architecture, since MIH is an IP-based solution that simply requires a thin client at the terminal and an MIH server in the IP backhaul, effectively future-proofing present networks.

Longer battery life, removing the need for performing periodic scanning of other radio access technologies at the mobile terminal, since MIH provides the availability of heterogeneous network services in a location-based manner.

No radio access network modifications, as the MIH client in the mobile node communicates with the MIH server for all local interconnections and handovers.

The degree under which the benefits of the MIH framework will vary depends on the regulatory scenario for using the TV white spaces. This will be analysed in subsection 5.10.8-.

5.10.5- Handover Mechanism

This section describes a COGEU based case for MIH based mobility management. It will be described as follows. First, there will be a specific use case development; based on the COGEU combined sensing and geo-location database scenario; as well as only geo-location database access scenario. Then an information flow will be given to show the stages of the diagram.

5.10.6- Scenario 1 for Commons: Sensing and geo-location database required

5.10.6.1 Scenario description Figure 46 shows a network topology for the commons scenario where sensing and geo-location database are required as described in Section 3.2-. In this section, mobility management features to support seamless handovers in case of the appearance of unregistered PMSE devices is considered. A mobility management strategy in this scenario is important both for the TV white space network as well as for the protection of the PMSE.


of 97

Figure 46: Network topology assuming sensing and geo-location database access requirements.

This scenario is motivated by the fact that, information on DVB-T incumbents as well as registered PMSEs, usually for professional applications, is stable; however, the unpredictability of unregistered PMSE applications and Electronic News Gathering, which requires protection, is the main challenge in the design of the mobility management features in the communication networks operating under the commons regulatory regime. In particular, the TV white space network has to move to other spectrum bands once an incumbent system is activated and detected, both to avoid harmful interference and degrade QoS performance in terms of error rate.

5.10.6.2 Information flow

The following steps gives details on the information flow for handover mechanism in a Master-Slave setting in the commons mode as shown in Figure 47.

1. Assume there is data exchange between Slave and Master Devices going on in a given channel, for example Channel one.

2. A PMSE device is activated, and starts transmitting in Channel one. If close to the Master-Slave, it will experience some interference.

3. A Slave Device performs sensing on Channel one, and detects PMSE activity. 4. After detecting incumbent activity, the Slave device sends a report to the Master device that

Channel one is busy (occupied by incumbent). 5. The Master Device then sends a command to the Slave device to pause transmission in

Channel one. 6. The Slave Device pauses transmitting in Channel one. 7. Data exchange between the Slave and Master Devices eventually stops in Channel one. This

step effectively protects the incumbent device from unduly interference. 8. The Master Device reports Channel one busy status to the Geo-Location Database and at the

same time requests for available channels list. 9. The Geo-Location Database updates Channel one to ‘occupied’ status and performs channel

availability search.


of 97

Figure 47: Information flow for hand over mechanism when a user discovers the operation of a PMSE device through sensing – no MIH function needed.

10. If the Geo-Location Database finds vacant channels, it provides the Master Device with the list

of available channels.


of 97

11. After receiving the list of available channels from the Geo-Location Database, the Master Device commands the slave device to sense if there is any activity in the given channels.

12. The Slave Device performs sensing in the available channels to detect any activity of incumbent (or other white space systems).

13. The Slave Device reports to the Master Device on channels with ‘NO PMSE activity detected’ status as candidates for handover.

14. The Master Device send an ACK signal to the Geo-Location Database indicating that it is going to use vacant channel(s). Notice that this implies the channels that can’t be used are occupied by incumbents.

15. The Geo-Location updates the channel availability status accordingly. 16. The Master Device commands the slave device to handover to the identified vacant channels. 17. The communication session between the Master and Slave Devices resumes in the vacant

white space channel(s).

In this case the MIH framework has not been used due to the simplicity of the scenario. Implementing the MIH for the purpose of handover from one channel to another would cause unnecessary system overload and decrease the overall efficiency. This is due to the fact that the device has to perform the actual sensing to determine if the bands are available or not – in this case assistance from the network is limited to the geo-location database. Future work will consider vertical handover under the joint sensing and geo-location database access.

5.10.7- Commons Scenario 2: Only geo-location required

5.10.7.1 Scenario description: Figure 48 gives a generic network topology for MIH based handover mechanism from LTE to WiFi operating on the TV white spaces assuming the commons with only geo-location database access required (the regulatory scenario is given in Section 3.3). In this section, mobility management features to support seamless handovers through the MIH framework is presented. A mobility management strategy in this scenario is important to realize the key benefits of exploiting the TV white space as presented before in the document. In this section, we focus on the handover mechanism of which is triggered by the mobile device. The network, through the 3rd Party Mobility Management Entity assists the hand over process from an LTE network connection to a WiFi access network through the MIH framework (and the ANDSF). In particular, we consider what happens when a handover event occurs. The handover procedure is triggered by MIH_LINK_SAP of the Slave Device, which sends an event to the MIHF_MN when it acquires information on the availability of WiFi networks operating on the TV white spaces in its vicinity, through WiFi beacons or from the MIIS server.


of 97

Figure 48: Network topology for MIH based handover mechanism from LTE to WiFi operating on the TV white spaces assuming the commons with only geo-location database access required.

5.10.7.2 Information flow

Figure 49 gives the information flow for handover from a serving LTE Master Device to a candidate WiFi Master Device. In this flow, it is assumed that the initial state of the Slave Device is a connection to an LTE Master Device. The Slave Device receives information on the availability of a WiFi network from the 3rd Party Mobility Management Entity, possibly the MIIS server or from beacons from the WiFi network and decides to switch to the WiFi network based on the user’s preference. The user may prefer a cheaper WiFi network to access the internet over the 3G network. The following is what happens when a handover occurs. A 3rd party mobility management entity (MIIS Server in this case) informs the User Device on available WiFi network on the TV white spaces. Based on the user preferences, a slave device decides to switch to available WiFi in the TV white spaces. The handover procedure is triggered by the MIH_LINK_SAP of the Slave Device which sends an event to the MIHF_MN (the MIHF in the slave device) indicating that the handover is needed. This procedure can also be triggered by a ‘link quality is degrading’ event. The handover trigger is conveyed by the MIHF_MN to the Serving LTE Master Device, which performs a database query on the MIIS to retrieve handover policy. The handover policy indicates the availability of the Candidate (target) WiFi Master Device, to which the Slave Device desires to gain access.


of 97

Figure 49: Information flow for user triggered – network assisted handover mechanism from LTE to WiFi operating on the TV white spaces under the commons mode implementing the MIH

function Then the MIHF_MN of the Slave Device tries to associate with the Candidate WiFi Master Device. When the association succeeds, the Slave Device notifies both the Serving LTE Master Device and the Candidate WiFi Master Device. The latter, also completes the handover procedures with network-to-network messages with the MIIS from which it learns from the outcome of its handover decisions. The detailed sequence diagram is illustrated in Figure 49, where the LTE Master Device is the Serving point of access and the WiFi Master Device is the target one. The specification of messages in the sequence may require vendor-specific extensions, which can be easily accommodated due to the flexibility of IEEE 802.21 messages. In the diagram some messages can happen simultaneously, and hence their order in the figure is arbitrary. In this scenario, the MIIS and the WiFi Master Device have access to the geo-location database, and the availability of the TV white spaces resources has not been explicitly indicated in the handover process. In this case, the impact of TV white space availability on the handover process is minimal.

5.10.8- Analysis of the impact of COGEU mobility management approaches to the system requirements

In this section, an analysis of the two mobility management approaches described in subsection 5.10.6- and 5.10.7- for the commons handover in the joint sensing and geo-location access regulatory scenario and only geo-location access regulatory scenario respectively is done. The analysis also includes two other regimes not studied in this work, but are used for control purposes. These are conventional (pure) sensing approach and Broker based secondary spectrum trading.


of 97

As can be seen in Table 6, the impact of conventional sensing based mobility management to the system requirements is mostly negative. This implies that sensing alone might not be the best way forward. However the requirement for Reconfigurability of the cognitive radio device for accessing spectrum and network services is similar in all situations. Reconfigurability is more a function of user preference than of resource shortage. In other aspects, pure sensing approach is quite subjective depending on extra factors such as hardware design, and intended usage environment, hence categorizes as neutral.

Table 6: Measuring spectrum mobility management approaches versus system requirements

Spectrum Mobility Management Under Different Scenarios

Co

nv

en

tio

na

l P

ure

Se

nsi

ng

Se

nsi

ng

&

G

eo

-L

oca

tio

n D

ata

ba

se

Acc

ess

(C

om

mo

ns

un

de

r R

eg

ula

tory

S

cen

ari

o 1

)

On

ly G

eo

-Lo

cati

on

A

cce

ss

(Co

mm

on

s u

nd

er

Re

gu

lato

ry

Sce

na

rio

2)

Bro

ke

r S

eco

nd

ary

S

pe

ctru

m T

rad

ing

(R

eg

ula

tory

S

cen

ari

o 2

)

Syste

m R

equ

irem

ents

Ability to protect incumbent systems -/+ -/+ + ++ Flexibility for operative coverage extension and preference networking - -/+ + ++ Support for QoS requirements of secondary user systems - -/+ + ++ Optimal choice of access technology -/+ + ++ + Requirement for geo-location database access - + ++ ++ Efficient use of the TV white spaces through secondary spectrum trading - -/+ + ++ Power saving for the end user device to acquire spectrum context - - + ++ Security vulnerability to attacks such as denial of service - -/+ + ++ Support for regulatory flexibility such as dynamic adjustment of operation power limits and frequency range

+ + ++ ++

Exploitation for context information such as location, network availability etc

+/- + + ++

Usability of the MIH framework for mobility management services - -/+ + ++ Cost efficiency and simplicity of network roll-out - - + ++ The need for autonomous sensing ++ + -/+ - Flexibility in spectrum usage through aggregation - -/+ + ++ Support for frequency planning - -/+ ++ + Reconfigurability of the cognitive radio device for accessing spectrum and network services

++ ++ ++ ++

Regulatory enforcement/changes -/+ + ++ + Support for system evolution -/+ -/+ ++ ++

Key: Positive force + ; Negative impact - ; Neutral effect +/- ; Strongly positive ++


of 97

From the COGEU perspective, the Sensing & Geo-Location Database Access (Commons under Regulatory Scenario 1) approach is a transitional approach. In this category, impact to power saving for the end user device to acquire spectrum context and cost efficiency and simplicity of network roll-out requirements have the lowest score due to the additional hardware and signalling overhead caused by the sensing requirement respectively. Otherwise, this scenario has a neutral impact to most system requirements because of its dependence on the activity level of incumbent systems in the geo-graphical location that the system operates. In general, Only Geo-Location Access (Regulatory Scenario 2) for both Commons and Broker based Secondary Spectrum Trading has a positive to strongly positive impact to the system requirements, respectively. This is mainly due to the centralized availability of TV white space information. Moreover, the Geo-Location Database supports both commons and secondary spectrum trading and hence has a very high impact to the support for regulatory flexibility such as dynamic adjustment of operation power limits and frequency range.

Therefore, mobility management under different regulatory assumptions scenarios will impact system requirements to different degrees. In that regard, the usability of the MIH framework for mobility management services varies in accordance to the regulatory scenario, as has been shown in the table. In this work, only the commons regime under both joint sensing & geo-location database access and only geo-location database access regimes have been presented.

5.10.9- Conclusion and future directions

This section has described the features for mobility management architecture to support seamless handovers in heterogeneous COGEU use cases over TV white spaces environments for commons regime under both joint sensing & geo-location database access and only geo-location database access approaches. In the first regulatory scenario, the MIH framework has not been used due to the simplicity of the scenario. Implementing the MIH for the purpose of handover from one channel to another would cause unnecessary system overload and decrease the overall efficiency. However, future work will consider vertical handover under the joint sensing and geo-location database access. In the second regulatory scenario, the MIH framework has been used to enable the handover from LTE network to a WiFi network through the support of the MIIS server. In this case, the impact of TV white space availability on the handover process was seen to be minimal due to the support of the geo-location database and the MIIS server. An analysis to measure the impact of spectrum mobility management approaches against system requirement has been performed. It was shown that mobility management approaches under different regulatory assumptions scenarios will impact system requirements to different degrees. In that regard, the usability of the MIH framework for mobility management services varies in accordance to the regulatory scenario. Future work will involve further analysis of various mobility management aspects in relation to system requirement and regulatory scenarios.


of 97

6- Instantiation of the COGEU reference architecture

COGEU three key application scenarios: WiFi with cognitive access to TVWS-spectrum commons; LTE extension over TVWS – secondary market; Public safety applications with cognitive access to TVWS – prioritization [D2.1].

6.1- LTE over TVWS

6.1.1- COGEU Instantiation of LTE over TVWS

The extension of LTE operational mode over TVWS is in line with what is being proposed by standardization bodies, namely ETSI RRS in TR 102 907 “Use Cases for Operation in White Space Frequency Bands” largely focus the use of LTE in the TVWS band, either in “Mid-/long range wireless access over white space frequency bands (with no/low/high mobility)” or “Sporadic use of TV white space frequency bands”. However, since white spaces are not a contiguous band, some modes can be better suitable than others. Up to now there are very few networks based on LTE. However, in the very near future this technology will be widely adopted. When it starts to being deployed as a commercial network the used frequency will be 2.6GHz in most European countries [17]. This high frequency can limit LTE and it is possible that in some areas the LTE’s coverage may be only characterised by some spots, not achieving the expected performance, and being capacity limited. As presented in [17], the excellent propagation characteristics of COGEU operation band, that starts from channel 40 to channel 60 (622MHz-790MHz),allows operators to cover higher geographical areas with less base stations (BSs), notably for a region of 624km2, LTE@700MHz only requires 4 BSs while [email protected] requires 80 BSs to cover the same area; which is less costs for operators. Furthermore, with a better signal-to-noise ratio (SNR), which is achieved with better propagation conditions in lower frequencies, the use of higher modulation schemes, that increases the system’s capacity, is possible. Thus, the temporary use of TVWS bands for LTE when extra capacity or higher coverage are required seems to be an efficient solution, namely on the real-time secondary spectrum market model. However, the use of LTE in TVWS shall fulfil some logical steps. As in any cellular network, operator does measurements regarding the network performance. This monitoring is done in order to assess the quality of the network (mostly, in terms of quality of service (QoS) and quality of experience (QoE)). The selected key performance indicators (KPIs), e.g., drop call rate, are useful to detect possible alarms in the network and trigger an action. For COGEU, this action is to perform a carrier request for a lower frequency (TVWS channels) to the COGEU Broker. Based on the same KPIs that triggered the carrier request or other parameters and taking into account the type of service, the number of carriers that are needed to resolve that particular situation is assessed. Along with the number of requested carriers, the operator’s network sends to the COGEU Broker at least the information for desired bandwidth, location for LTE BS and associated coverage and the estimation for service’s duration. With the received request, the broker uses the GPS address of the location or its coordinates to inquire the Geolocation Spectrum Database in order to check if in that location there are free TVWS bands that may provide the requested carriers and desired bandwidth. Moreover, the broker also needs to check inside its modules the following parameters: if in that location the possible free bands are already occupied with secondary users – TVWS Occupancy Repository –, the trading information for the leasing of the spectrum and coexistence policy information with other bands / cells / technologies – in Trading Information Repository and Policies Repository, respectively. If everything was ok, the final consumer (which in cellular networks is the operator), shall be informed about the available carriers and bandwidth, time availability, leasing price, maximum allowed transmit power (see D6.1 for more details). Finally, the operator’s network shall inform the broker if it accepts or rejects the offer. In case of rejection, the process shall restart. In case of acceptation, and depending on leasing contract, the bands are delivered to the operator (that later will be assigned to users) and the TVWS Occupancy Repository is updated.


of 97

The process discussed above is summarized on Table 7. The number of phases depends on the granularity for the whole process.

Table 7: Generic phases for requesting and acquiring TVWS carriers

Phase Logical steps

Network monitoring

Normal process in any cellular network

Detection of a high variation in a defined KPI (lack of coverage or capacity)

Send a request for TVWS carriers

Carrier request Use of technology particularities, e.g., for LTE: time x frequency grid, AMC according to SNR, different bandwidths, fixed RRB size (12 sub-carriers, 15 kHz each and 0.5 ms of a slot duration), and

Estimation of number of carriers, necessary bandwidth and duration is performed

Information send to COGEU Broker

COGEU Broker inquiry

Inquiries in primary and secondary users allocation databases, and if no user is found

Checks trading information and coexistence policy

Information send to network operator

Carrier allocation Requires acceptance from the operator, otherwise the process shall restart

Update of TVWS Occupancy Repository

With LTE, new demands on the Operation and Maintenance (O & M) of the network are foreseeable. Therefore, in addition to evolving existing management solutions, E-UTRAN/EPC also encompasses some new Self-Organized Network (SON) functionalities such as Auto-Configuration, Auto-Optimization and Self-Healing:

o Self-Configuration of Base Stations (BSs) will reduce the amount of manual processes involved in the planning, integration and configuration of new Base Station. This will result in a faster network deployment and reduced costs for the operator in addition to a more integral inventory management system that is less prone to human error.

o The Auto-Optimization has as objective to maximize network performance, optimizing the configuration while taking into account regional characteristics of radio propagation, traffic and UE mobility in the service area is effective.

o For Self-healing has as its objective when some nodes in the network becomes in

operations, self-healing mechanisms aims at reducing the impacts from the failure, for example by adjusting parameters and algorithms in adjacent cells so that other nodes can support the users that were supported by the failing node.

In COGEU the introduction of SON can minimize the operation costs of running a network by reducing and eliminating manual configuration of network operational parameters at the time of network planning, network deployment, network operations, and network optimization while fitting into the existing operational processes and procedures that are currently in place today. The previously described process is illustrated for the LTE case in Figure 50.


of 97

Figure 50 : Possible process to request and acquire TVWS carriers for LTE

The previous process, i.e. the process to request and acquire a TVWS carrier, can be described as follows:

1) Manual configuration and carrier allocation is done during the planning phase. 2) The information about the carrier that was allocated is sent to the specific Base Station (BS). 3) Normal communication process between BS and User Equipment (UE) is done and also the

information about the carrier. 4) The UE and BS measurements are sent to Core Network (CN). 5) In order to optimise the network (i.e. interference, capacity or coverage), new carriers are

requested to COGEU system; the location, required bandwidth, and duration is also sent. 6) The CN is informed about the allocated carriers, and that information is also sent to BS (2). 7) The TVWS Occupancy Repository is updated with the channels that were accepted by the

operator and became in use. Through the solution’s ability to configure, optimize and recover automatically, the LTE SON will offer operators operational cost savings associated with network planning, network deployment and network optimization. Next the characteristics of FDD and TDD are discussed in order to have a better insight of which is better suitable to operate in TVWS.

6.1.2- FDD and TDD

The convergence of voice, video and other data services is the main goal of many service providers. To achieve this, technologies associated with traditional voice services are being replaced with technologies capable of providing bandwidth demands of today’s consumer. LTE is a strong candidate to take the leadership on this field and since it is a flexible technology it can both satisfy user’s demands and operator’s deployment constraints (deployment costs, interfere, etc). The good propagation conditions of UHF bands (where TV white spaces are located) is seen as an opportunity for operators to reduce their deployment and operational costs while providing the same services or even better services to users with a less cost per bit.


of 97

For LTE two duplex modes are considered: Frequency Division Duplex (FDD) and Time Division Duplex (TDD). FDD, which historically has been used in voice-only applications, supports two-way radio communication by using two distinct radio channels. Alternatively, TDD uses a single frequency to transmit signals in both the downstream and upstream directions. FDD uses a pair of frequencies, one frequency for upstream transmissions and the other frequency for downstream transmissions. With pairing frequencies, simultaneous transmission in both directions is possible. To mitigate self-interference between upstream and downstream transmissions, a minimum amount of frequency separation must be maintained between the frequency pair, a guard band. TDD uses a single frequency channel to transmit signals in both the downstream and upstream directions. It operates by toggling transmission directions over a time interval and thus a guard time (instead of a guard band) between transmit and receive streams is required.

6.1.2.1 Data symmetry FDD systems utilize channel plans that are comprised of frequencies with equal bandwidth. Since each channel has a fixed bandwidth, the channel capacity of each frequency also is fixed and equal to that of all other channels in the frequency band. This makes FDD ideal for symmetrical communication applications in which the same or similar information flows in both directions, such as voice communications. In TDD the toggling between downlink and uplink directions takes place very rapidly and is imperceptible to the user. Thus, TDD can support voice and other symmetrical communication services as well as asymmetric data services. TDD also can handle a dynamic mix of both traffic types. The relative capacity of the downstream and upstream links can be altered in favour of one direction over the other. This is accomplished by giving a greater time allocation through time slots to downstream transmission intervals than upstream. This asymmetry is useful for communication processes characterized by unbalanced information flow, such as Internet. FDD can also be used for asymmetric traffic. However, in order to be spectrally efficient, the downstream and upstream channel bandwidths must be matched precisely to the asymmetry. Since Internet traffic is bursty by nature and the asymmetry is always changing, the channel bandwidth cannot be precisely set in FDD. In this respect, TDD is more efficient.

6.1.2.2 Spectrum Efficiency Frequency spectrum is a scarce need. This scarcity drives the need to optimize the use of available bandwidth. FDD systems operate on the principle of paired frequencies. FDD channel plans maintain a guard band between the downstream and upstream channels. The guard band is required to avoid self-interference between both channels and, since it is unused, essentially is wasted spectrum. In contrast, TDD systems require a guard time (instead of a guard band) between transmit and receive streams. The Tx/Rx Transition Gap (TTG) is a gap between downstream transmission and the upstream transmission. This gap allows time for the base station to switch from transmit mode to receive mode and subscribers to switch from receive mode to transmit mode. During this gap, the base station and subscriber are not transmitting but are simply allowing the base station transmitter carrier to ramp down, the Tx/Rx antenna switch to actuate, and the base station receiver section to activate.

6.1.2.3 Conclusions The discussion above has highlighted some differences between TDD and FDD, which are both supported by LTE. Considering the extension of LTE over TVWS the question shall be focused more on the efficient use of the spectrum. In this sense FDD requires two symmetric bands (which is not always easy to find), one for upstream and other for downstream plus a guard band between both directions and for adjacent channels. TDD only requires one carrier that is shared in a time basis which eliminates the need for a guard band between both directions. Other advantage of TDD mode is the fact that dynamic bandwidth allocation may be done between the downlink and uplink, which is very suitable for asymmetric traffic and can also be used in symmetric traffic. However, FDD is easier to implement because there is no need for time synchronism and transmissions in both directions may happen at the same time.


of 97

6.2- WIFI over TVWS

The wide spread of Wi-Fi technology is apparent in all environments. Extremely high numbers of domestic and private networks exist in all possible locations, especially in dense populated areas. The affordability and ease of deployment of a Wi-Fi network has made it also an attractive option for larger public networks known as hotspots. Furthermore a number of operators use Wi-Fi for blanket coverage of large areas and offer their customers Wi-Fi connectivity and mobility in large areas where they operate in. Under these developments the use of Wi-Fi in the COGEU regulatory scenarios, can vary. The attractive characteristics of TVWS bands in conjunction with the characteristics of Wi-Fi technology can be a powerful enabling technology for the commercial success of TVWS usage. According to the section 2-, the IEEE 802.11af TG is working towards the proposal and standardization of an 802.11 variation for use in the TVWS bands. The characteristics of the technology, which will be able to exploit the lower bands, will include the support of single or multiple usages of channels if available. The data rate of the devices will depend on the availability of the channels in the specified areas. Also the devices will be able to operate in contiguous and non-contiguous channels in order to be able to utilise the empty spaces on TVWS. 802.11af will be able to operate in 5, 10 and 20 MHz channel width. OFDM will be used and fixed subcarrier spacing is recommended in order to limit complexity. In line with the work carried out in IEEE 802.11af TG the characteristics of the TVWS bands require spectrum saving practices, this is important since the first goal of the WSD deployments is the protection of the incumbent systems, namely DVB-T and PMSE. Due to the fact that a WiFi can be operated in a 5 MHz channel, this option has been selected as an important requirement for the Wi-Fi over TVWS deployment. Use of Wi-Fi over TVWS using 5MHz bandwidth allows for a protection level due to the fact that the spectrum affected by the Wi-Fi transmission is more limited than the normal operation of the Wi-Fi which uses 20MHz. It is possible to use one DVB-T channel only to deploy a Wi-Fi network. Although the adjacent channels are affected there is a difference of 20 db, from the peak for a measured channel (see Figure 51), in accordance with Wi-Fi spectrum mask. Conversely, if you narrow the available bandwidth to a narrower set of frequencies, you have less noise and increase your SNR, thus you increase your range. The fact is that although the maximum TX rate of the Wi-Fi Network will be limited to 16.25Mbps the range and the quality of received signal will be better than using a 20MHz channel. Operation on first regulatory scenario In the case where the Wi-Fi is operated under the first regulatory scenario where the operation is under the spectrum of commons with sensing and geo-location database access, Wi-Fi technology will be deployed in infrastructure mode, using a central access point. This access point will be the gateway and control centre for clients accessing resources outside the network and internet. Access to the geo-location database will only be possible from the access point which will download the operation instructions (channel, power), based on the requests, and configures the clients accordingly. Access point will also be responsible for the compliance of the clients to the regulations and instruction from the geo-location database and any regulatory authority. The access point can limit the range of the coverage and also will be able to control whether a client can transmit or not. The model of master-slave will be applied on the Wi-Fi operation; in this case the master is the access point and the slaves are the devices that are attached to the access point. Regarding the sensing capabilities of the Access Point, it is required that the AP will be able to sense, either periodically or on request. This is useful in case that the processing and resources of an AP are limited and also to limit the overhead if there is no need for sensing operation in the operated area. The sensing data will be transmitted to the geo-location database for consideration when the AP requests a renewal or initial operation instruction for a specific area.


of 97

Figure 51: Measured Wi-Fi 5MHz channel The operation of the Wi-Fi AP in the spectrum commons with sensing capabilities can be as follow:

AP power on

Location determination

Contact DB providing location

If requested start initial sensing operation

Transmit sensing data to DB

Receive operation instructions for available channels and maximum power

AP Configures slave devices

AP starts periodic sensing operation if requested

If all OK start transmission to slave devices

Slave devices start transmitting using the provided frequencies

Operation of second regulatory scenario In the case of second regulatory scenario (see Figure 52 : Wi-Fi over TVWS in the Spectrum Commons mode in both regulatory scenarios), where spectrum commons and secondary spectrum market operates with only the geo-location database access as required, Wi-Fi can be used in both operation modes. It can also be used as a deployment of a network by acquiring a frequency band from the broker. The difficulty on this case is that because Wi-Fi need almost three channels to operate as it does in the 2.4 GHz bands it will be more expensive to operate. Because the decision for using the Wi-Fi on 5MHz channel limits the amount of required channels from at least 3 to 1, it makes it a compelling option for use under the broker architecture and as one of the available technologies for licensing control operation. It can be a viable and attractive option the deployment of Wi-Fi in TVWS band with stable paid frequencies, also due to the lower infrastructure cost and very high availability and familiarity between the possible clients and perspective users. In the second scenario Wi-Fi can also operate under the spectrum of commons case with similar configuration with the first regulatory scenario. The difference is that it will not need to have sensing capabilities, which it makes it easier and with less overhead to process.


of 97

The operation of the access point in the spectrum of commons for the second regulatory scenario, without sensing capabilities are identical to the AP with sensing capabilities but without the sensing parts.

Figure 52 : Wi-Fi over TVWS in the Spectrum Commons mode in both regulatory scenarios

6.3- Public safety over TVWS

As introduced in previous tasks T3.1 we consider Use of Dynamic Spectrum Access Radios by the Public Safety Community. There may be considerable opportunity for dynamic spectrum access radios to be used by the public safety community and within public safety frequency bands. We also note the potential for reconfigurable radios to alleviate many of the interoperability issues associated with public safety spectrum use. Manage interoperability Public Safety domain is characterized by many different wireless heterogeneous networks like TETRA, TETRAPOL, Analogy Professional Mobile Radio (PMR) and satellite communications. In some cases, commercial systems like GSM/GPRS are used. In large disasters, military entities might operate together with Public Safety organizations. As a consequence, there is an issue of interoperability when an emergency crisis is to be resolved by different public safety organizations equipped with different communication systems. Coupled with cognitive radio systems operating in the TVWS, centralised spectrum management with geolocation as in COGEU, can be a technology enabler to resolve the interoperability barriers at technical level by activating the needed waveforms on the cognitive radio platform. The spectrum regime considered could be spectrum common, or secondary spectrum trading. In this later case Public Safety agencies might enter the market as secondary spectrum players. For the sake of simplicity, we’ll refer to both geo-location database or COGEU broker in the next paragraph as “database”. In a typical scenario for this use case, the Public Safety user has a non-fixed (portable or mobile) device and is riding in a vehicle (or device is embedded in the vehicle). The user wants to have connectivity to another device which is also moving. Typical deployment scenarios include urban areas and rural areas where the user may connect to other users in peer-to-peer or ad-hoc networks. This deployment


of 97

scenario is typically characterized by a master device with low antenna height, internet connectivity by some connection that does not utilize TV white space, and some means to predict its path of mobility. This knowledge of mobility could be simple (GPS plus accelerometer), sophisticated (GPS plus routing and mapping function) or completely specified by the user via user-interface. A simplified operational scenario utilizing TV white space to provide peer-to-peer connectivity service in a mobility environment consists of the following steps:

1. The mobile master device powers up with its TVWS radio in idle or listen mode only (no active transmission on the TVWS frequency band).

2. The mobile master has internet connectivity and establishes a connection to a trusted white space database.

3. The mobile master registers its current geolocation, address contact information, etc. associated with the owner/operator of the master with the trusted database administrator (if not currently registered).

4. Following the registration process, the mobile master will send a query to the trusted database requesting a list of available WS channels based upon its current location and a prediction of its future location, extrapolated from the motion or mobility of the device. The current location is specified in latitude and longitude. The method to specify the future location is TBD, potential methods include movement vector (direction and velocity), a set of latitude/longitude points which specify a closed polygon where the future location is within the polygon, or similar.

5. If the mobile master has been previously authenticated, the database responds with a list of available white space channels that the mobile master may use, and optional information which may include (1) a duration of time for the use of each channel (2) a maximum transmit power for each channel.

6. Once the mobile master authenticates the WS channel list response message from the database, the master selects an available WS channel from the list for use.

7. The other user device in the peer-to-peer connection scans the TV bands to locate a mobile master transmission, and associates with the mobile master. The slave/user device queries the master for a channel list, based on the slave's device identification, geolocation and optionally a prediction of its future location.

8. If required by local regulation, the master device verifies the slave's device identification with the database.

9. If allowed by local regulation (e.g. the slave's device identification is verified by the database), the mobile master provides the list of channels locally available to the slave/user device. If the channel that the slave/user terminal is currently using is not included in the list of locally available channels, the slave/user device ceases all operation on its current channel. The slave/user device may scan for another Master's transmission on a different channel.

Nevertheless, operational requirements for communication systems in the Public Safety domain are usually different from the Commercial domain especially in terms of reliability, availability, responsiveness and security. Those specific requirements could be addressed through differentiated spectrum access where Public Safety systems could be given better priority. Emergency Situation and Disaster Relief involving Access priority and Pre-emption for Public Safety organisation Organizations involved in handling emergency operations often have a fully owned and controlled infrastructure, with dedicated spectrum, for day to day operation. However, lessons learned from recent disasters show such infrastructures are often highly affected by the disaster itself. To set up a replacement quickly, there is a need for fast reallocation of spectrum, where in certain cases spectrum can be freed for disaster relief. To utilize free or freed TVWS spectrum quickly and reliable, automation of allocation, assignment and configuration could be performed thanks to the COGEU Broker. This approach does in no way imply such organizations for disaster relief must compete on spectrum allocation with other white spaces users, but they can. Service priority can be introduced into the COGEU broker, with each secondary system having an assigned priority level. In general, channel availability for equal priority services is determined based on the trading mechanisms described in section 5.1.2-. In this manner, secondary TVWS systems will avoid selecting channels that are already in use by other secondary systems, enforcing coexistence. The concept of service priority can be a complementary solution to allow some systems (e.g., emergency public safety systems) to operate with higher priority over other services. For example, the


of 97

database can provide differing channel lists (and/or other radio operating parameters, such as maximum allowed transmit power levels) based on service priority. In order to react to an emergency incident, the broker may allocate channels to public safety systems that were previously allocated to other lower priority services or simply prioritize access to them during the allocation of frequencies. When the public safety systems arrive on a scene, it would access the broker, and the broker would have the option of assigning channels that are currently utilized by lower priority systems. The lower priority systems would also receive a new set of channels/radio operating parameters for the new spectrum usage scenario. While priority based broker should preferably be operated on a near real-time basis, even hourly broker updates could potentially provide much more efficient usage of available spectrum. Another possibility is to re-open the market (and related auction and trading mechanism) when a major crisis happens. All players re-enter the auction game but Public Safety systems have a better chance to win due to higher priority. The following figure Figure 53 is illustrated the concept of spectrum access priority. Player 1 and Player 2 are pre-empted from their original spectrum in order to give acces to emergency responder in case of critical disaster relief. This scenario is only possible when national regulation policies are implemented for this specific and extraordinary situation.

Figure 53 : Illustration of Spectrum priority access for Emergency responders In Figure 53 the Fire Brigade is illustrated as using ad-hoc networking. This is given as an example illustrating the specific case where all nodes are master nodes in a way that they allocate TVWS channels from the COGEU Broker. However, a backhaul link may not be available to all nodes, such as depicted for the 3 PMR terminal nodes in the far right of Figure 53. To handle TVWS channel allocation for such nodes, a master node (the Fire Truck in the figure) with a backhaul link relays or proxies the Broker query for them. So, nodes without a backhaul link follow the procedure as defined for clients. The ad hoc network radios utilise the provided TVWS channels. Details on forming and maintenance of the ad hoc network, including repair of segmented networks caused by segments operating on different RF channels, is out of scope of spectrum allocation.


of 97

7- Conclusions and future work

The main task during this second part of WP3 was to derive a stable version of the COGEU reference architecture, and define the basic functional blocks and their interoperations to implement a real time spectrum management platform for both Spectrum Commons and Secondary Spectrum Market. This work relates to COGEU tasks T3.3 “COGEU Reference Architecture for Spectrum Commons “and T3.4 “COGEU Reference Architecture for Secondary Spectrum Market”. The work reported in this D3.3 Deliverable has been done while closely following the demonstrator implementation performed under WP7. Indeed, during demonstrator preparation partners took a special care of clearly highlighting what are the key features they will be focusing on (see section 3.1-) , while using the reference architecture as a basis for implementation. Moreover, on the many features and requirements extracted from the uses-cases, not all of them are currently being implemented in the demonstrator. A prioritization process has been followed in order to focus on the most promising features and the most essential requirements. Figure 54 shows the interconnection between the work presented in this deliverable and further COGEU tasks.

Figure 54: Main interconnections between D3.3 and further tasks of the project

Now that WP3 is over, and the final COGEU reference architecture detailed in this document, the consortium is confident to consider that the originally defined WP3 objectives have all been fulfilled:

A set of system level requirements required to design future cognitive radio-mobile systems have been extracted from the COGEU use-cases. (reported in D3.1)

Investigation regarding algorithms for detecting TVWS and for supporting their exploitation is achieved (reported in D3.1).

Definition of the COGEU reference architecture based on the system level requirements for the spectrum commons and the secondary spectrum market model is delivered in this document.


of 97

8- References

[1] M. Mueck, K. Sithamparanathan, et. al., "ETSI RRS - The Standardization Path to Next Generation Cognitive Radio Systems", CogCloud wokshop, collocated with the IEEE Personal Indoor and Mobile Radio Communications (PIMRC) 2010, Istambul, Turkey, September 2010.

[2] OECD (2005), “Secondary Markets for Spectrum: Policy Issues,” OECD Digital Economy Papers, No. 95, OECD Publishing. doi: 10.1787/232354100386

[3] Patrick Xavier, Dimitri Ypsilanti, “Policy issues in spectrum trading,” Info, Vol. 8, No. 2, pp. 34 – 61, 2006

[4] Monisha Ghosh, Vasanth Gaddam and Kiran Challapali.“Proposed modification to Section 8.4 of Draft ECC Report,” Philips Research North America. New York

[5] FP7 ICT-2009.1.1 COGEU. D3.1 “Use-cases Analysis and TVWS System Requirements,” August 2010

[6] Junjik Bae; Beigman, E.; Berry, R.; Honig, M.L.; Hongxia Shen; Vohra, R.; Hang Zhou; , "Spectrum Markets for Wireless Services," . 3rd IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks, 2008. DySPAN 2008, vol., no., pp.1-10, Oct. 2008

[7] Berry, R.; Honig, M.L.; Vohra, R.; "Spectrum markets: motivation, challenges, and implications," IEEE Communications Magazine, vol.48, no.11, pp.146-155, November 2010

[8] D. Ardagna, C. Batini, M. Comerio, Marco Comuzzi, F. De Paoli, S. Graga, B. Pernici, “Negotiation Protocols Definition,” Multichannel Adaptive Information Systems (MAIS), Nov, 2004

[9] Mark Bykowsky, A secondary market for the trading of spectrum: promoting market liquidity, Telecommunications Policy, Volume 27, Issue 7, Practical Steps to Spectrum Markets, Pages 533-541, August 2003

[10] John W. Mayo, Scott Wallsten, Enabling efficient wireless communications: The role of secondary spectrum markets, Information Economics and Policy, Volume 22, Issue 1, Wireless Technologies, March 2010, Pages 61-72

[11] Nguyen Hoang, Thuan. Types of electronic payment system: The requirements from different actors’ perspective (Part 2) [Internet]. Version 5. Knol. April 2009 2. Available from: http://knol.google.com/k/thuan-nguyen-hoang/types-of-electronic-payment-system-the/25lgke3rt3f2g/7.

[12] Lehr, W. and N. Jesuale (2008) "Public Safety Radios Need to Pool Spectrum," IEEE Communications Magazine, March 2009.

[13] Michael J. Barclay 1, Terrence Hendershott ,“A comparison of trading and non-trading mechanisms for price discovery”, 18 March 2008, pp 846

[14] Patroklos Argyroudis, Timothy Forde, Linda Doyle and Dona lO'Mahony, “A policy-driven trading framework for market-based spectrum assignment”, 2007

[15] M. Kumar and S. I. Feldman. "Internet Auctions", in Proceedings of the 3rd USENIX Workshop on Electronic Commerce, Pages 49-60, August 1998.

http://knol.google.com/k/thuan-nguyen-hoang/types-of-electronic-payment-system-the/25lgke3rt3f2g/7

http://knol.google.com/k/thuan-nguyen-hoang/types-of-electronic-payment-system-the/25lgke3rt3f2g/7


of 97

[16] A Statement on Spectrum Trading: Implementation in 2004 and beyond, Ofcom, (2004)

[17] Everything Everywhere Limited, Response to the Ofcom consultation. “Reserving the band 2500-2690MHz for the London 2012 Games”. 17 September 2010. Source: http://stakeholders.ofcom.org.uk/binaries/consultations/band-2500-2690-london-2012-games/responses/Everything_Everywhere.pdf

[18] COGEU FP7 ICT-2009.1.1. D3.1 “Use-cases Analysis and TVWS Systems Requirements”. 2 August 2010

[19] ETSI TR 102 683 : Reconfigurable Radio Systems (RRS) ; Cognitive Pilot Channel (CPC)

[20] White Space Database Group ex parte submission(ET Docket No 04-186)

[21] IEEE 802.22 WRAN (doc IEEE 802.22-10/0073r03)

[22] Baykas, T. et al., Developing a Standard for TV White Space Coexistence: Technical Challenges and Solution Approaches

[23] OFCOM : Implementing geolocation: Publication date 9 November 2010. Available at http://stakeholders.ofcom.org.uk/binaries/consultations/geolocation/summary/geolocation.pdf

[24] Analysys, DotEcon, H&H, Study on conditions and options in introducing secondary trading of radio spectrum in the European Community, final report for the European Commission, Available: http://ec.europa.eu/information_society/policy/ecomm/radio_spectrum/_document_storage/studies/secondary_trading/secontrad_final.pdf, 2004

[25] Jon M. Peha, "Sharing Spectrum through Spectrum Policy Reform and Cognitive Radio," Proceedings of the IEEE, 2008

[26] Chen Guo, Tao Peng , Shaoyi Xu Cooperative Spectrum Sensing with Cluster-Based Architecture in Cognitive Radio Networks

[27] I. F. Akyildiz, W. Y. Lee, M. C. Vuran, and S. Mohanty, “Next Generation/ Dynamic Spectrum Access/ Cognitive Radio Wireless Networks: A Survey, ” Computer Networks Journal (Elsevier), Vol. 50, pp, 2127 – 2159, September 2006

[28] F. Granelli, P. Pawelczak, R. V. Prasad, K.P. Subbalakshmi, R. Chandramouli, J.A. Hoffmeyer, S. Berger, “Standardization and Research in Cognitive and Dynamic Spectrum Access Networks: IEEE SCC41 Efforts and Other Activities,” IEEE Communications Magazine, vol. 48, no. 1, pp. 71-79, Jan. 2010

[29] R. V Prasad, P. Pawelczak, J. A. Hoffmeyer, H. S. Berger, "Cognitive Functionality in Next Generation Wireless Networks: Standardization Efforts", IEEE Communications Magazine, vol. 46, no. 4, pp. 72-78, Apr. 2008

[30] K. Taniuchi and et al, “IEEE 802.21: Media independent handover: Features, applicability, and realization,” IEEE Commun. Mag., 47(1):112 –120, January 2009

[31] R. Bolla, R. Rapuzzi, and M. Repetto, “Handling mobility over the network,” In CFI ’09, pages 16–19. ACM, 2009.

http://stakeholders.ofcom.org.uk/binaries/consultations/band-2500-2690-london-2012-games/responses/Everything_Everywhere.pdf

http://stakeholders.ofcom.org.uk/binaries/consultations/band-2500-2690-london-2012-games/responses/Everything_Everywhere.pdf

http://stakeholders.ofcom.org.uk/binaries/consultations/geolocation/summary/geolocation.pdf

http://stakeholders.ofcom.org.uk/binaries/consultations/geolocation/summary/geolocation.pdf


of 97

[32] R. Ferrus, O. Sallent, and R. Agusti, “Interworking in heterogeneous wireless networks: Comprehensive framework and future trends,” IEEE Wireless Commun., 17(2):22 –31, April 2010

[33] IEEE, “IEEE Standard for Local and metropolitan area networks– Part 21: Media Independent Handover Services,” IEEE Std 802.21–2008, January 2009.

[34] InterDigital, “Media Independent Handover: White Paper,” April, 2009. Available: http://www.interdigital.com/images/id_pubs/InterDigitalMIHWhitePaper_Apr09.pdf

[35] J. Mwangoka, P. Marques, J. Rodriguez, “Cognitive Mobility Management in Heterogeneous Networks,” 8th ACM International Symposium on Mobility Management and Wireless Access (MOBIWAC 2010), 37 – 44, October 2010.

[36] Steven S. Skiena, “The Algorithm Design Manual”, Second Edition, Springer, ISBN: 978-1-84800-069-8

[37] Newcom++ project, DR9.1 “Identification of relevant scenarios, use cases and initial studies on JRRM and ASM strategies”

[38] Steven S. Skiena, “The Algorithm Design Manual”, Second Edition, Springer, ISBN: 978-1-84800-069-8

[39] Measuring Fragmentation of Two-Dimensional Resources Applied to Advance Reservation Grid Scheduling, Julius Gehr, Jorge Schneider, 9th IEEE/ACM International Symposium on Cluster Computing and the Grid

[40] Protocol to query a White Space Database draft-caufield-paws-protocol-for-tvws-01.txt available at http://tools.ietf.org/id/draft-caufield-paws-protocol-for-tvws-01.txt

http://www.interdigital.com/images/id_pubs/InterDigitalMIHWhitePaper_Apr09.pdf

http://tools.ietf.org/id/draft-caufield-paws-protocol-for-tvws-01.txt


of 97

List of Figures

Figure 1: Combination of unlicensed access to TVWS with secondary spectrum trading ..................................... 3 Figure 2 : White Spaces Network ......................................................................................................................... 8 Figure 3: QoSMOS Concept illustrated .............................................................................................................. 15 Figure 4 : SACRA ICT technical approach ........................................................................................................... 16 Figure 5: Spectrum manager interfaces ( Source: [21]) ...................................................................................... 18 Figure 6: Unlicensed Operation in the TV Broadcast Bands (Source: [20]) ......................................................... 18 Figure 7: ETSI RRS Database Access (Source: [19]) ............................................................................................. 19 Figure 8: 802.19 co-existence framework (Source: [22]) .................................................................................... 20 Figure 9: Initial COGEU frame structure [Source: D3.1]. .................................................................................... 24 Figure 10 : Reference Model for spectrum commons with sensing and geo-location access required ............... 26 Figure 11 : Geo-Location Database Data Structure ............................................................................................ 27 Figure 12: COGEU interfaces with the geo-location database [Source: D4.1] .................................................... 28 Figure 13 : Cooperative Spectrum Sensing in CR Networks CR 1 is shadowed over the reporting channel and CR

3 is shadowed over the sensing channel. ................................................................................................... 31 Figure 14: General architecture for combining geolocation and sensing information ........................................ 35 Figure 15: Geolocation database with allowing spectrum commons and secondary spectrum trading operations

.................................................................................................................................................................. 37 Figure 16: COGEU reference architecture for commons and secondary trading, only geo-location access

required..................................................................................................................................................... 37 Figure 17 : COGEU spectrum trading demonstrator (Source COGEU WP7 D7.1) ................................................ 38 Figure 18: Message Sequence Chart for COGEU Reference Architecture ........................................................... 39 Figure 19: The three phases for the allocation and trading of the TVWS as well as the maintenance of the

broker repository ...................................................................................................................................... 40 Figure 20 : TVWS allocation and trading mechanism based on D6.1 .................................................................. 41 Figure 21: TVWS trading mechanism based on D6.1 .......................................................................................... 42 Figure 22: Thick versus thin market ................................................................................................................... 43 Figure 23: Pricing mode protocol sequence diagram ......................................................................................... 44 Figure 24: English Auction Protocol ................................................................................................................... 45 Figure 25: Paypal transaction protocol (Paypal) (* Note that before send payment information, Paypal will

adjust the buyer and merchant’s Paypal account ) (source [11] ) ............................................................... 46 Figure 26: TVWS Occupancy Repository ............................................................................................................ 49 Figure 27 : Broker TVWS occupancy repository database structure. .................................................................. 50 Figure 28: Policies Repository functional block ................................................................................................. 51 Figure 29: Policy management tool ................................................................................................................... 52 Figure 30: Hierarchy of polices .......................................................................................................................... 54 Figure 31: Definition of the contract between buyers and banks ...................................................................... 55 Figure 32: Definition of the parameters to seller ............................................................................................... 55 Figure 33: Definition of the parameters to buyer .............................................................................................. 56 Figure 34 : Geo-location database access domains. ........................................................................................... 58 Figure 35: Example of list of available channels format reported by the geo-location database for a specific

location ..................................................................................................................................................... 59 Figure 36: The process of TVWS computation. .................................................................................................. 60 Figure 37 : COGEU Web-service ......................................................................................................................... 61 Figure 38 : Data Model Proposal ....................................................................................................................... 62 Figure 39 : Location Element ............................................................................................................................. 63 Figure 40: Authentication and Authorization Entities in COGEU system ............................................................ 64 Figure 41: Example of signing a message ........................................................................................................... 66 Figure 42: Possible scenario of hand-over across heterogeneous networks in TV white spaces ........................ 67 Figure 43: Functional components for spectrum context aware mobility management .................................... 69 Figure 44: MIH Reference model and SAPs ........................................................................................................ 71 Figure 45: Architecture for integrating the Media-Independent Handover framework with COGEU user cases

and the mobility management component ............................................................................................... 72 Figure 46: Network topology assuming sensing and geo-location database access requirements. .................... 74 Figure 47: Information flow for hand over mechanism when a user discovers the operation of a PMSE device

through sensing – no MIH function needed. .............................................................................................. 75


of 97

Figure 48: Network topology for MIH based handover mechanism from LTE to WiFi operating on the TV white spaces assuming the commons with only geo-location database access required. .................................... 77

Figure 49: Information flow for user triggered – network assisted handover mechanism from LTE to WiFi operating on the TV white spaces under the commons mode implementing the MIH function ................. 78

Figure 50 : Possible process to request and acquire TVWS carriers for LTE ........................................................ 83 Figure 51: Measured Wi-Fi 5MHz channel ......................................................................................................... 86 Figure 52 : Wi-Fi over TVWS in the Spectrum Commons mode in both regulatory scenarios ............................. 87 Figure 53 : Illustration of Spectrum priority access for Emergency responders .................................................. 89 Figure 54: Main interconnections between D3.2 and further tasks of the project ............................................. 90


of 97

List of Tables

Table 1: Channel availability based on geolocation-only information ............................................................ 33 Table 2: Channel availability based on sensing-only information .................................................................. 33 Table 3: Channel availability based on geo-location and sensing information ............................................... 33 Table 4: Final decision process for channel availability based on geo-location and sensing information ...... 34 Table 5: Relevant actions for combining geolocation and sensing information .............................................. 35 Table 6: Measuring spectrum mobility management approaches versus system requirements .................... 79 Table 7: Generic phases for requesting and acquiring TVWS carriers .......................................................... 82


of 97

List of Abbreviations

3GPP 3rd Generation Partnership Project

CEPT Conference of European Postal & Telecommunications

CR Cognitive Radio

DVB-T Digital Video Broadcasting - Terrestrial

DTV Digital Television

ETSI European Telecommunication Standards Institute

EU European Union

ENG Electronic News Gathering

FCC Federal Communications Commission

GSM Groupe Spécial Mobile (also, Global System for Mobile communication)

IEEE The Institute of Electrical and Electronics Engineers

ICT Information and Communications Technologies

IMT International Mobile Telecommunications

ISM Industrial Scientific and Medical (band)

ITU International Telecommunication Union

LAN Local Area Network

LTE Long Term Evolution

MAC Medium Access Control

MIMO Multiple-Input Multiple-Output

OFCOM Office of Communications

OFDM Orthogonal Frequency Division Multiplexing

PAWS Protocols to Access White Space database

PMSE Programme Making and Special Events

PU Primary User

QoS Quality of Service

R&D Research and Development

RF Radio Frequency

RRM Radio Resource Management

SAP Services Ancillary to Programme making

SDR Software Defined Radio

TV Television

TVWS TV White Spaces

UHF Ultra High Frequency

UMTS Universal Mobile Telecommunications System

US Unites States of America

VHF Very High Frequency

WiMAX Worldwide Interoperability for Microwave Access

WiFi Wireless Fidelity (IEEE 802.11)

WLAN Wireless Local Area Network

WP Work Package

final architecture for tvws spectrum sharing systems - cogeu

Documents