CHAPTER 1

INTRODUCTION

Internet traffic is due to the manifold effects of increasing number of broadband

subscribers and increasing bandwidth per subscriber. For an example from Japan, the

number of asymmetric digital subscriber line (ADSL) subscribers has begun to

decrease while the number of fibre to the home (FTTH) subscribers has exceeded 16

million .Such a trend of increasing bandwidth per subscriber naturally leads to the

increasing use of video-related services, such as YouTube. The average data size per

video content is also increasing because the increased bandwidth per subscriber

allows easy transmissions of higher definition videos.

The quality, or data size, of video is strongly influenced by the television (TV)

standard. Fig. 1 plots the necessary bit rate versus the number of pixels. The definition

of TV used to be the analog “National Television System Committee (NTSC)” or the

digital “standard definition TV” (SDTV), and is the “high-definition TV” (HDTV)

now. A higher definition TV, the so-called digital cinema, (4-k D-cinema), has

Already been standardized and commercialized in the movie industry. Beyond D-

cinema, the Japan Broadcasting Corporation (NHK) is promoting the research and

development (R&D) of the So-called “ultra high-definition TV” (UHDTV, 8-k). Even

after UHDTV, a high-definition 3-D TV would be yet to follow. Fig.2 summarizes the

UHDTV technology, where it describes well how immersive the images by UHDTV

are. Of course, most of data distributed to and/or exchanged among mass consumers

shall be more or less “compressed” such that the necessary Band width per user may

be much lower than the values plotted in Fig. 1. Generally speaking, however, the rate

of the traffic increase will follow the trend plotted in Fig. 1 even with advanced

compression technology. Fig. 3 is the envisioned time schedule of the R & D

activities for UHDTV conducted at NHK. The standardization for UHDTV is already

en route at relevant standardization bodies, such as Society of Motion Picture and

Television Engineers (SMTPE) and International Telecommunication Union-Radio

communication Sector (ITU-R). The commercial service shall be feasible around

2025.

The roadmap of the TV definition is also the roadmap of the bandwidth required to

transmit the uncompressed video signals .As the technology of TVs develops, the

transmission rate has to be correspondingly faster. As a consequence of the evolution

toward high definition, the means of broadcasting needs also change from the

traditional terrestrial TV with a limited bandwidth to other means with a much higher

bandwidth, such as A novel cable TV technology. Even for satellites, broadcasting

many channels of UHDTV would be quite challenging even with advanced

compression technologies. Not only for such technical reasons, will the forthcoming

telecom-broadcasting convergence also lead TV broadcasting to be through networks.

Likewise, it is of critical importance to develop high-capacity network technologies

suitable for providing services with massive high-definition video data, including

future TV programs, in order to perpetuate the traffic growth.

The telecom-broadcasting convergence with a high-capacity network will

also bring about the emergence of new applications, such as a high-definition version

of YouTube-like services and high-definition teleconferencing or immersive

telepresence services. In fact, the forecast traffic increase is mostly due to the

increasing video traffic, as plotted in Fig. 4. According to this survey, the most of

traffic will contain either video or peer-to-peer (P2P) shared data, both of which are

huge file Transfer. Other statistics regarding the average Internet traffic in Japan

Show a compound annual growth ratio (CAGR) of approximately 40%, as plotted in

Fig. 5. It also plots the two-survey results of the total power consumption of the IP

routers in Japan, conducted by Japanese Government in 2001 and 2006, respectively.

In fact, the power consumption of the IP routers in Japan had jumped by ten times

from 2001 to 2006, while the traffic growth during this period had been almost

comparable. Incidentally, the power consumption discussed here does not include the

power for cooling. The correlation between the traffic and the power consumption of

the IP routers is inevitable because the power consumption of an IP router depends on

the power consumption of the forwarding engine and is therefore almost proportional

to the throughput. On the other hand, the total annual power generation of Japan is

almost constant around 1000 TWh, which means that approximately 1% of the total

power supply was consumed by the IP routers in Japan in 2006. Although the portion

is still merely a few percents at present, considering the proportionality between the

traffic and power consumption, the traffic growth at a CAGR of 40%, sustainable for

the next few decades, would require unlimited use of power even exceeding the total

power supply of Japan, as extrapolated in Fig. 5. This merely states that the today’s IP

technology is incapable of scaling the capacity to the growing traffic. We also note

2

that not only the power, but also the footprint of an IP router will become too large to

scale.

FIG-1

(Trend of TV technology. The bandwidth required to transmit the uncompressed contents versus the number of pixels)

FIG. 2(UHDTV in contrast to HDTV)

3

FIG -3

(Research and development timeline for UHDTV along with pertinent standardization events.)

We first discuss our proposed dynamic optical-path network (DOPN) where the

extremely energy-efficient feature of optical switches will be discussed, followed by

the discussions on the intrinsic differences between the packet and circuit switching,

and then, the scalability of the DOPN will be explored along with the key photonic

technologies. Then it will describe one of our specific efforts toward the realization

of the DOPN, i.e., the development of the technology for transmitting UHDTV video

signals using integratable optical time-division multiplexing (OTDM) technology,

over what we call “ultrafast all-optical LAN–SAN (storage area network),” a

miniature version of the DOPN. The power consumption targets for the OTDM

devices will also be discussed briefly.

4

FIG-4

(Projected compositions of Internet traffic)

FIG-5

(Total power consumption of the IP routers and the traffic growth)

5

CHAPTER- 2

LITERATURE REVIEW

2.1 DYNAMIC OPTICAL-PATH NETWORK

A. Extremely Low-Energy Potential of Optical Switches

The observations in the previous section call for a highly efficient network technology

that handles a few orders of magnitude higher capacity while consuming even less

power than present .Such a dramatic improvement cannot be realized through

incremental processes, but a “clean-slate” approach. To the best of authors’

knowledge, only the optical switch has such a potential.

The potential of optical switches is that the power consumption does not depend on

the line rate, but only on the port count. Therefore, at a certain break point of the line

rate, the optical switch will become more energy efficient than the electronic switch,

and the effectiveness will be more conspicuous as the traffic grows. For example, an

80 × 80 micro electromechanical systems (MEMS) switch is reported to consume

approximately 22 W, which means a power efficiency of 0.275 W/port, or an energy

efficiency of 2.75 PJ/bit for a line rate of 100 GB/s. In contrast, the Cisco’s CRS-3

router is allegedly reported to have 6 NJ/bit energy efficiency [8]. Here, we already

find three orders of magnitude difference. Let us suppose a hypothesis those optical

switches to be used instead of IP routers, as illustrated in Fig. 6, then all the switching

functions shall be done by the optical switches—a network in which optical switches

are mutually connected in a mesh configuration, dynamically provisioning end-to-end

connections. If this were feasible at all, the energy consumption would be much

smaller. We propose to call such networks “DOPNs.” This of course requires an

additional control plane network, but considering the necessary capacity for the

control plane network, its energy consumption should be negligible as compared with

that for the data plane network for high-definition video contents Veeraraghavan et al.

Proposed a circuit-switched high speed end-to-end transport architecture called

“CHEETAH,” in which a packet-switched network is used as the primary network,

While a huge file is transferred via a secondary optical-path network. The

challengesin their proposal are mainly how to establish an end-to-end optical paths

6

across metropolitan area network (MAN) and WAN and how to determine which

network to be used in order to best utilize the overall network whose parameters are

always varying.

7

Fig-6

(Proposed hypothesis to use an optical switch with

an amplifier per port instead of an IP router)

On the other hand, one of the most challenging technical breakthroughs necessary for

the realization of DOPNs is nothing but the technology of high-port-count optical

matrix switches. Thus far, a number of R&D efforts have been made to realize a

8

practical optical switch with high port count. The technology available in market at

present is the MEMS switch technology. This technology has already been widely

used in service for reconfigurable optical add drop multiplexers (ROADMs) and

optical cross connects (OXCs). While the feature of high port count is promising by

MEMS switches, the cost for inspection may be an issue for the application to DOPN

where a large number of low-cost optical switches have to be deployed. On the other

hand, silicon photonics has many intrinsic physical properties that are attractive for

the application to high-port-count optical matrix switches, i.e., small chip size,

efficient and fast thermo-optic effect, potentially low cost in fabrication and

inspection, etc. In fact, a simple design estimation shows that a 256 × 256 optical

matrix switch based on thermo-optic silicon photonics could be only 5×5 cm2 large

and consume only 20mWper port , or an energy efficiency of 0.2 pJ/bit for a line rate

of 100Gb/s, which is even one tenth of the MEMS switch. Fig. 7 plots the power

consumption versus throughput for the CRS-3 class IP router and the silicon-

photonics-based switch with a port count of 128 × 128. We assume that the optical

amplifier consumes 3 W per port, and hence, the power consumption of an optical

switch including optical amplifiers becomes 3.02 W per port in total. Even though the

power for optical amplification is predominant, the power consumption of the optical

node can be a few orders of magnitude lower for higher throughputs, as indicated in

Fig. 7.

Fig-7

9

(Power consumption versus throughput)

B. Circuit Switching Versus Packet Switching

Apparently, we have not discussed the impact of optical switches instead of

conventional IP routers so far. The Monrovian discussions on the fundamental

difference between the IP-router-based and optical-switch-based network can be

found. According to these studies, the elementary bottlenecks in packet and circuit

switching can be evaluated in terms of the mean file transfer time, and analytically

described with M/G/1-processor sharing (PrS) and M/G/1 first-in first-out (FIFO)

models, respectively. Let us define the pertinent parameters as follows: S is the file

size, E[S] is the mean file size, C is the transmission speed, λ is the frequency of

requests of file transfer, Var[S] is the variance of the file size, and G is the guard time

for the optical circuit provisioning. We ignore the guard time for the packet switching

(PS) as well as any influences from transmission control protocol (TCP) congestion

controls. Therefore, the PS considered here is ideal. The aforementioned analysis

suggests that the circuit switching better suit the applications in which the transferred

file sizes do not vary much. As the context of this paper, we argue that such

applications would be video-related services, like video on demand (VOD) and/or

future high-definition YouTube-like services. Also, an important feature of the circuit

switching is the intrinsically perfect Quos once the circuit is established. In this sense,

it is also suitable for teleconference services. In general, video-related services are

better served through circuit-switched networks, as summarized in Table I. Also

technically speaking, it is important to note that it is far much easier to realize a high-

bandwidth circuit-switched network than an equivalently high-bandwidth packet-

switched one by making use of optical communications technology. A hybridization

of modest-speed PS and high-speed OCS has been proposed. The point of proposal

there was compelling, and in fact, the effectiveness of the hybridization has been

clearly demonstrated through a fundamental simulation. Fig. 8 depicts the image of

the hybridization, where an end host always starts sending and/or receiving data

through the PS network while simultaneously requesting an optical path to OCS. And

once the optical path is established, the data transfer port is immediately switched

10

from the PS network to OCS. The results are plotted in Fig. 10, whose details have

been reported.

Fig-8

2.2 ROUTER THROUGHPUT BOTTTLENECK

Figure 9 depicts expected traffic volume increase and ICT power consumption

increase in Japan over the near term twenty years period. Compared to 2006 levels,

traffic volume and ICT power consumption at 2025 are expected to increase by 190

times and 5.2 times, respectively. The traffic increase rate assumed was 32 %, which

is smaller than that of about 40 % as presently measured in Japan. The power

consumption estimation incorporates reasonable CMOS technology advances, but will

reach 20% of Japan’s electric power generation capability in 2025. Router power

consumption is among the key factors driving the power consumption increase.

Figure 10 depicts recent advances in core router throughput. The chart clearly

indicates that the throughput advances appear to be saturating, which stems from the

power consumption limitations of LSIs. The fall in CMOS driving voltage has

recently saturated and leakage current increases substantially as gate length decreases.

In order to create transport networks, routing/switching on a lower layer than layer 3

IP routing offers better power efficiencies and so the throughput can be enhanced at

the cost of coarse switching granularity Among the lower layer transport mechanisms,

the optical path/waveband cross-connect provides the highest efficiency. Routing

granularity is arbitrary with IP/MPLS routers and Ethernet switches etc., while other

lower layer systems offer fixed granularity. It is therefore reasonable to use the best

combination of different layer technologies.

11

Fig-9

(Estimated traffic increase and power consumption of ICT)

Fig-10

(Advances in core router throughput)

2.3 TRANSPORT ARCHITECTURE OF THE FUTURE

12

2.3.1 Impact of Video

Highly granular routing/switching is very effective in collecting relatively small

capacity data streams. For example, when sensors become ubiquitously distributed

around the globe, say, 10 times the world’s population, the IP-based Internet

mechanism works well and will be indispensable in collecting such relatively small

streams that are spatially distributed. The collected/aggregated data should be

transported in the network with the lowest layer transport technologies possible,

instead of hop-by-hop IP routing. Optical paths have thus been initially utilized in the

network to cut-through routers (Fig. 11 middle). In future networks, video-oriented

traffic is expected to be dominant. Progress in high-definition and ultra-high-

definition TV (more than 33M pixels) is steadily advancing, and the expected source

video bit rate will reach 72 GB/s per channel. The inefficiencies of the present IP

protocol will become more evident given the advances in video-oriented services.

Bandwidth demanding applications such as ultra-high definition video will directly

use optical paths/circuits, as illustrated in Fig. 11.Regarding traffic volume, in sensor

networks, even if each sensor produces 1 kbps and the number of sensors is 7 billion,

then the total generated bit rate around the globe is just 7 Tera bits per second. Please

note that this is equivalent to just 1,000 ultra-high definition (72 Gbps) video channels

(Fig. 12). The impact of broadband video is thus significant, and so will be a major

factor in designing future networks.

Fig-11(Electrical router cut-thorough and optical fast circuit switching)

13

Fig. 12(Different bit-rate services and the bandwidth)

2.4 Hierarchical Optical Path –Introduction of Wavebands

In terms of power efficiency and throughput, lower layer switching is more efficient;

however, the flexible bandwidth path capability provided by LSPs can be more

efficient than the more rigid bandwidth path capabilities enforced by lower layer

switching. Therefore, TDM paths such as VCs (Virtual Containers) in SDH and

ODUs in OTN (Optical Transport Network) are hierarchically structured as shown in

Fig. 13; the lower order paths provide service access, while the higher order paths

generally provide transmission access. At present, a wavelength path (channel) is

defined and utilized as a single order entity. As traffic demand and fibre transmission

capacity increases, much larger bandwidth optical paths, the waveband, will be

introduced. When optical layer services such as OVPN (Optical Virtual Private

Network) services, lambda leased line services, optical circuit (circuit and path are

used interchangeably in this paper) or optical flow switching services emerge, the

hierarchical optical path arrangement will be needed. Optical fast circuit switching

14

will be suitable for creating a nation-wide super-high definition source video

distribution network that connects video headed nodes.

Fig-13(SDH, OTN, and OP architectures and path capacities)

15

Fig-14(Wavebands and hierarchical optical cross-connect)

2.5 Benefits of Wavebands

An optical switch can switch multiple optical paths. Switching groups of optical paths

or wavebands can reduce the total switch size (necessary number of cross-connect

switch ports) substantially. This mitigates one of the major challenges; the need to

create extremely large scale optical cross-connects. For example, when the waveband

add/drop ratio is less than 0.5, switch scale reductions of more than 50% for a matrix-

switch-based cross-connect system , and more than 20% for a WSS/WBSS

(Wavelength/Waveband Selective Switch) based cross-connect system have been

confirmed. The role of wavebands in realizing efficient optical circuit switching

networks has been clarified. Figure 15 depicts optical path establishment in a single

optical path layer network as well as that in a multilayer optical path network. In a

single layer optical path network, optical path establishment/tear-down requires node

(optical cross-connect) by node optical switch setting. On the other hand, in a

multilayer optical path network, optical path establishment can be done utilizing one

(direct) or multiple wavebands. As a result, in the connection establishment/release

phase, the number of nodes involved in the signalling process is greatly reduced and

the connection set-up/release delay is minimized. The relationship between the optical

wavelength path cross-connect and the waveband cross-connect corresponds to that of

the electrical switching system and the cross-connect system in POTS networks. With

Regard to connection establishment and control/signalling, traffic-driven (optical flow

switching) or control driven (optical circuit switching), and centralized or distributed

control scheme can be applied as demanded by networking requirements.

16

Fig. 15

(a) Single layer optical path network (b) Hierarchical optical path

network

(Comparison of single layer optical path and hierarchical

optical path networks)

2.6 WOBAN

Hybrid wireless-optical broadband access network (WOBAN) is emerging as a

promising technology to provide economical and scalable broadband Internet access.

In this cross-domain network architecture, end-users receive broadband services

through a wireless mesh front-end which is connected to the optical backhaul via

gateway nodes. In this article, we present the architecture and functional

characteristics of a WOBAN prototype built in the Networks Lab. at UC Davis. We

cite some research challenges on hybrid networks based on our experimental

17

observations. WOBAN shows excellent promise for future access networks. This

cross-domain network architecture consists of an optical backhaul (e.g., a Passive

Optical Network (PON)) and wireless access in the front-end (e.g., WiFi and/or

WiMAX). In WOBAN, a PON segment starts from the telecom Central Office (CO)

with an Optical Line Terminal (OLT) at its head end. Each OLT can drive several

Optical Network Units (ONU), and each ONU can support several wireless routers of

the wireless frontend in WOBAN. The wireless routers directly connected to the

ONUs are called as wireless gateways. The wireless front-end also consists of other

wireless routers to provide end-user connectivity. Therefore, the front-end of a

WOBAN is effectively a multi-hop Wireless Mesh Network (WMN) which is

connected to the high-capacity PON segment in the back-end, creating cross-domain

integrated network architecture.

2.6.1 WOBAN ARCHITECTURE

18

Fig-16

Figure 16 shows the architecture of WOBAN prototype developed in the Networks

Research Laboratory at UC Davis. The wireless routers form the WOBAN front-end

and connect to the end users (who can be scattered over the geographic area served by

the WOBAN and who are not shown in Fig. 1). These wireless routers support data

rates up to 54 Mbps. Several designated routers are configured to have Gateway

capabilities (by loading appropriate open source firmware) and each such Gateway is

connected to an ONU via a 10/100 Base-T Ethernet port. The wireless routers are

placed with an effective distance of 50-60 meter between pairs. Two OLTs (Optical

19

Line Terminal) emulate the functionality of the telecom Central Office (CO) of the

general WOBAN architecture. Each OLT can drive several ONUs using an optical

splitter. The OLTs and ONUs are connected through Ethernet PON (EPON) ports.

The OLTs are connected to the Rest of the Internet (ROI) using the campus-wide

backbone network at UC Davis. The prototype architecture is divided into three

planes: (a) Control Plane, (b) Data Plane, and (c) Management Plane. The Control

Plane is used to define different control features of the nodes in the WOBAN

prototype. The Data Plane configures routing and different data transfer scenarios, and

collects measurement data for different experiments. The Management Plane is used

For remote access and programmability of the prototype nodes. The WOBAN

Network Operations Centre (NOC) (see Fig. 16) is responsible for the management of

all these planes.

2.6.2 Distinguishing Features

The WOBAN prototype has several distinguishing features which are different from

other related prototypes reported in the literature, as follows. To the best of our

knowledge, this is the most integrated wireless-optical hybrid network test bed. Other

Test beds have only a small number of nodes and have been used as proof of

concepts. On the other hand, WOBAN prototype features programmability, self

organization, and slice-based experimentation. The WOBAN prototype is large

enough to demonstrate its useful properties, e.g., two OLTs can demonstrate fault-

tolerance properties of WOBAN so that, if one OLT breaks, the other parts of the

WOBAN can “self organize” themselves to still carry the affected traffic through the

other operational parts of the WOBAN. The self-organization property of WOBAN

also holds for (1) other failure types, e.g., ONU failure, fibre cut, wireless router

failure, etc. and (2) optimal routing. The deployment and management cost of

WOBAN prototype is low as it is built from highly-customized off-the-shelf

components, open sources, and indigenous software. The front-end can be set up as a

plug-and-play wireless mesh. The prototype nodes feature programmability. The

open source firmware provides the programmability in the wireless routers. The

programmability of OLT can be performed by using the craft port in the OLT box and

the ONU programmability can be emulated by gluing a separate “Linux box” with

20

each ONU. The prototype is reconfigurable and provides self organizing and self-

healing properties. The reconfigurability is performed by Layer-2 (L2) connectivity

And intelligent routing. Power consumption of the wireless nodes is very low (1-2.5

watts/router). As the wireless mesh constitutes a large part of the prototype, the

overall power consumption is also low.

CHAPTER-3

CONCLUSION

21

Energy efficiency in telecom networks is gaining significant attention among the

telecom networks researchers. In this dissertation, we developed novel methods and

techniques to build energy-efficient next generation telecom networks. The

algorithms, architectures, design methods, and results presented in the dissertation

will assist researchers and telecom service providers in developing networks in an

energy-efficient manner. In this chapter, we summarize the important contributions

and findings in the dissertation.

We showed how to build a prototype for a novel, high-bandwidth future

access network technology, named WOBAN. This technology is envisioned to satisfy

future bandwidth demand of technology-savvy customers in a cost-effective manner,

and it can be an attractive solution for future “last-mile” access networks. We

demonstrated the performance of several typical applications such as data transfer,

voice, and video over our WOBAN prototype. We observed that too many wireless

Hops degrade the application performance, particularly for video. Future research

challenges accumulated from our prototyping experiences were also illustrated. The

WOBAN prototype will be instrumental to develop, test, and analyze the performance

of hybrid network protocols. This programmable and configurable access architecture

Will facilitate future experimental, hybrid, and cross domain networking research.

REFERENCES

1. R. Tucker, “Optical packet-switched WDM networks: A cost and energy

perspective,”

22

2. Ieee journal of selected topics in quantum electronics, vol. 17, no. 2,

march/april 2011 by Shu Namiki, Member, IEEE, Takayuki Kurosu, Ken

Tanizawa, Member, IEEE, Junya Kurumida, Member, IEEE,Toshifumi

Hasama, Hiroshi Ishikawa, Fellow, IEEE, Tsuyoshi Nakatogawa, Madoka

Nakamura,and Kimiyuki Yamaha

3. Hybrid Wireless-Optical Broadband Access Network (WOBAN):Prototype D

evelopment and Research Challenges Pulak Chowdhury, Suman Sarkar, Glen

Kramer, Sudhir Dixit, and Biswanath Mukherjee

4. Optical Technologies that Enable Green Networks Ken-ichi Sato,

Fellow,IEEE Nagoya University, 464-8603 Furo-cho, Chikusa-ku, Nagoya,

Japan

5. http://www.wikipedia.org/

23

24

25