Artificial Intelligence, Visual Communications & Networks (AVCN) Research Centre

uws.ac.uk

Muhammad Zeeshan Shakir

School of Engineering and Computing

05/02/2018 1

Artificial Intelligence, Visual Communications & Networks (AVCN) Research Centre

• Strong and active research group• 09 Academics • 8 postdoctoral research fellows +

KTP associates• 30+ PhD students• Active collaborations

• Main research areas• Wireless/mobile/5G networking• Condition monitoring• Event/Task Scheduling and

Optimization • Logistics management and

Immersive technology • Intelligent decision support

systems • Big data, IoTs, Cloud computing• Video & image compression &

transmission• Connected and remote health

Over 700 peer-reviewed publications; over 200 publications since 2012, with 15 best

paper awards since 2012

05/02/2018 2

5G technologies and Mobile cloud

computing

Unmanned aerial vehicles,

Wireless/mobile networking, and Image processing

Internet-of-Things and Intelligent systems and

AI

Embedded Systems and HPC, and

Digital Signal Processing

Smart &Sustainable Green

Technologies

EU H2020 project “SliceNet– End-to-End Cognitive Network Slicing and Slice Management Framework in Virtualised Multi-Domain, Multi-Tenant 5G Networks”, funded with Euro 700k for UWS activities, 2017-2020. https://5g-ppp.eu/slicenet/

EU H2020 project “SELFNET – Framework for Self-organized Network Management in Virtualized and Software Defined Networks”, funded with Euro 700k for UWS activities, 2015-2018. https://selfnet-5g.eu/

UWS project “Future connectivity for rural and sparsely populated areas by exploiting Network Flying Platforms with millimetre-wave, and massive Multiple Input Multiple output implementation”, partner with BT, NATS, T4i Engineering , £96k for UWS postgraduate, 2017-2020.

EU Erasmus Mundus project “SmartLink”, AVCN is the project coordinator for the consortium of 25 EU and South Asian university partners, funded with Euro 695k for UWS activities, 2014-2018.

EU Erasmus Mundus project “gLINK”, AVCN is a partner in the consortium of 18 partners, Euro 246k for UWS activities, 2014-2018.

Digital Health Institute project “Smartcough”, funded with £90k (2015-2017). AVCN leading a consortium formed by UWS, University of Edinburgh, Cirrus Logic (former Wolfson Microelectronics), and CHSS.

UK MOD project “Introducing Disruptive Technologies to Upgrade Military Visual Communication Capacity”, 2015.

UKRC project “A Pilot Study on a Fully Deployable Cooperative Unmanned Aerial Vehicles System for Flooding Prediction”, 2014-2015.

Royal Society & NSFC project “Research on Multiple UAV Cooperation for Marine Oil Spill Detection”, 2014-2016.

Royal Society of Edinburgh “Hyperspectral Imaging Vehicle-based Epidemiology (HIVE): Improving early warning systems for infectious disease outbreaks funded with £4200, 2016-2017.

CENSIS and Thales Optronics Ltd. project “Thales-Challenge: Low-pixel Automatic Target Detection and Recognition (ATD/ATR)” £87k, 2015-2016.

KTP project, “Eye Tracking: Predicting a viewer’s likely response to digital advert” with Lumen, £240k, 2016-2019.

Industrial project “Emotional Gym: Interactive Bike, Lorretto housing”, Wheatley Group, £50k, 2016-2017.

KTP project “Service Oriented Web Applications Development”, 2014 -2016.

KTP project “EVIVA: Evaluating the Education of Interpreters and their clients Through Virtual Learning Activities”, 2014-2015.

KTP project “GPU implementation of HEVC video processing”, 2014-2016.

Industrial project “Service-Oriented Cloud Information Centre for Airport Services”, 2014-2017.

British Council project “Emerging 5G Technology for Disaster Management” (supported workshop in Pakistan).

QNRF project “Low Power Reconfigurable self-calibrated Multi-Sensing Platform For Gas Applications”, 2013-2015.

Funded 5G-related Projects

05/02/2018 3

Networked-Flying Platforms:Paving the Way Towards B5G Wireless Networks

Agenda

• Introduction to 5G systems – Motivation to future systems: densification – Wireless fronthaul: Role of Network flying platforms

• Vertical FSO-based framework for fronthaul – Link budget, weather conditions and system losses– Performance evaluation in terms of data rate– Cost evaluation (capital) for various technologies

• Drone-small cell association problem– Problem formulation and numerical analysis – Two algorithms: distributed and central

– Airborne SON– Layered architecture for cellular network– Drone placement problem and SON

– 3-Takeaway Points

05/02/2018 5

Nikola Tesla(10 July 1856 – 7 January 1943)

“When wireless is perfectly applied, the whole earth will be converted into a huge brain, which in fact it is, all things being particles of a real and rhythmic whole. We shall be able to communicate with one another instantly, irrespective of distance.” Nikola Tesla (1925)

05/02/2018 6

http://www.teslasociety.com/biography.htm

Roadmap to 1000x: sustainable way!

Advanced

Interference

coordination

Self

Organizing

Networks

Hyper-dense HetNets

Ad-hoc Small-cells and other

radio access technologies

Wi-Fi Integration

2020: Beyond 4G, Radio Evolution for the Gigabit Experience, Nokia Siemens Networks, Aug. 2011

• Logical separation between idle mode and userDTX/DRX

• More antenna >> higher rate >>more time forsleep

• Higher capacitywithout desired

densification

Milli-meter and • Microwave wave

spectrumVisible Light

spectrum

Unlicensed

spectrum(ASA)

Spectrum aggregation• Phantom cell >>

control/data planedecoupled

• Cognitive principles in spectrum and resource allocation

• Trafficoffload• Improved and reliable

links

• Green backhaul solutions forhyper dense small cells

• D2D communicationsand other device

centric services

• Load balancing• CoMP and other

networkcoordination >> energy savings

05/02/2018 7

Network Densification: Pitfall

Source: Project: Post card from the near future, CTVR , Ireland, 2014.

05/02/2018 8

Backhaul/fronthaul Dynamics: Ultra-dense HetNet

Internet

Backhaul Hub

RRUFronthaul

Lightly loaded residential area during peak time

Core network

Central controller

• RF NLOS backhaul/fronthaul • Placement of backhaul/fronthaul hubs (NP-Hard

problem)• Data rate of Gbits/s could not be achievable

Fiber backhaul

• Fiber • Deployment cost • Latency due to bending, etc.

Wireless fronthaul

Free space optical (FSO) communications: towards the speed of wireline links . . .

Capacity routing fronthaul link

Heavily loaded business area during peak time

Fronthaul over shared/adjacent channel

Multi hope links

05/02/2018 9

Real World Scenario: Wireless NLOS Backhaul

05/02/2018 10

http://www.blinqnetworks.com/

Drones for Cellular Technology: Two sides of the same coin

• Tower maintenance & repair• Inspection and construction strategy• Safety and operational efficiency

05/02/2018 11

05/02/2018 12

Vertical FSO-based Fronthaul for Ultra-dense HetNets

M. Alzenad, M. Z Shakir, H. Yanikomeroglu, and M.-S. Alouini, “FSO-based vertical fronthaul/backhaul framework for 5G+ wireless systems,” in IEEE Communs. Mag., vol. 56, no. 1, pp. 218-224, Jan 2018, https://arxiv.org/abs/1607.01472

Spectrum for Vertical NFP-based Communications

Technology Data rate Channel modeling Atmosphere Wind

Free space optics (FSO):

• Up 10+ Gbits/sec• LOS isrequired• Unlicensed• Reduced interference

• Link budget analysis• model forFSO

satelliteComm• mathematical

models

• Fogand visibility• Cloudthickness

• Combinationof

weather

• Wind shear and wind speeddislocate the UAV

• May lead to higher scintillation errordue to UAV motion

RF (sub-6 GHZ)

• Mbits/sec• LOS/NLOS• Licensedand costlyspectrum• Higherinterference under NLOS

• Somesuburban,rural models

• FSPL+ NLOS/LOS

• Limitedmodelsover subGHz and5GHz

• Negligible atmosphere

• Atmospheric particle absorptionis very low overthe lowerelectromagneticspectrum

• Negligible signal fluctuationsas compared with fading dueto obstacles

• Signal fluctuationsdue to wind, vegetation,etc.

Mm-wave/Terahertz

• Few Gbits/secto10x Gbits/sec• LOS• Unlicensed/lightly licensed• Reduced interference –massive

MIMO

• No vertical models• Trails ongoing

• Rain attenuation• Doppler effect

05/02/2018 13

Data Rate of Vertical FSO Links

05/02/2018 14

Data rate and Link Margin under Weather Conditions

05/02/2018 15

data rate Link Margin

Worst Weather and Variable Divergence Angle

05/02/2018 16

data rate Link Margin

Favourable Solutions: Vertical Fronthaul under Different Weather

Weatherconditions FavourableSolutions

• Adaptive transmit power control• Low altitude flights• Systems parameters such as Divergence angle• Hybrid FSO/RF/mm-wave system

• Low or extreme high altitude flights (less than 5 km or more than 20 km (higher clouds are very thin))

• Site diversity (via multi hop communications between hubs)

• Hybrid FSO/RFsystem• Sitediversity• Low altitude flights• Systems parameters such as Divergence angle

05/02/2018 17

Total Cost of Ownership: Comparison for various Backhaul/Fronthaul

A snapshot of the typical Poisson distributed HetNet Deployment cost of several backhaul/fronthaul solutions

05/02/2018 18

NFP-Small cell Association Problem Formulation

05/02/2018 19

S. A. W. Shah, T. Khattab, M. Z. Shakir, M. O. Hasna, “A distributed approach for networked flying platform association with

small cells in 5G+ Networks,” in Proc. IEEE GLOBECOM’2017, Singapore, Dec. 2017. https://arxiv.org/pdf/1705.03304.pdf

Distributed Maximal Cells Algorithm (DMCA)

05/02/2018 20

Simulations: NFP-SC Association

05/02/2018 21

Airborne Self Organising Networks A-SON

05/02/2018 22

H. Ahmadi, K. Katzis, M. Z. Shakir, “A novel Airborne self-organising architecture for 5G+ networks,” in Proc. IEEE VTC-

Fall’2017, Toronto, Canada, Sep. 2017. https://arxiv.org/pdf/1707.04669.pdf

Functionalities of Layers

● Higher Layer (HL): NFPs operating in the HL belong to High Altitude Platform (HAP) category and are responsible for optimizing the resources in transport networks for lower layers.

● Medium Layer (ML): NFPs in the ML belong to the Medium Altitude Platform (MAP) category and are responsible for relaying the network between the lower and higher layer. NFPs in the medium layer are dual role playing i.e., in addition to relaying, MAPs are performing surveillance to ensure safe and secure operation of the architectures.

● Lower Layer (LL): NFPs operating in the LL are typical low altitude platforms (LAPs) flying at relatively lower altitudes and responsible for network optimization, NFP placement and providing capacity and expanding coverage.

05/02/2018 23

Spoil for Choices over Platform Selection?

05/02/2018 24

3-Takeaway Points – it’s a blue-sky

➢ Self-organization in Airborne network is an open research framework, e.g., placement and distribution of NFPs across the layers to achieve latency, massive connectivity or extreme broadband by exploiting cross-layer optimisation frameworks.

➢ Brownfield solution: NFPs can join the key players of 5G communications for immediate impact of the technology with right combination, e.g., Hybrid FSO/mm-wave medium should be considered as a potential solution to meet the demands of 5G networks and emerging NFP use cases could accelerate B5G applications.

➢ Interdisciplinary research efforts are required to make the story successful. e.g., wind turbulence, regulatory alignments, security/privacy, operational and capital expenses are to bedone.

05/02/2018 25

Flying Platforms for 5G and B5G: Some Ongoing Developments

• Sep 2017: BT&EE – Drone-way between Isle of Lewis and Mainland.

• May 2017: Qualcomm - LTE unmanned aircraft systems

• Feb 2016: Intel testingdrones over AT&T LTE Networks, Verizon starts 5G Trials with Samsung

• Jan 2016: Project Skybender: Google's secretive 5G internet drone tests revealed

05/02/2018 26

Recent References: NFPs plus B5G

NFPs for Wireless fronthaul

1) M. Alzenad, M. Z. Shakir, H. Yanikomeroglu, M.-S. Alouini, “FSO-based vertical backhaul/fronthaul framework for 5G+ wireless networks,” IEEE Communs. Mag., vol. 56, no. 1, pp. 218-224, Jan 2018. https://arxiv.org/pdf/1607.01472.pdf

2) S. A. W. Shah, T. Khattab, M. Z. Shakir, M. O. Hasna, “A distributed approach for networked flying platform association with small cells in 5G+ Networks,” in Proc. IEEE GLOBECOM’2017, Singapore, Dec. 2017. https://arxiv.org/pdf/1705.03304.pdf

3) S. A. W. Shah, T. Khattab, M. Z. Shakir, M. O. Hasna, “Association of networked flying platforms with small cells for network centric 5G+ C-RAN,” in Proc. IEEE PIMRC’2017, Montreal, Canada, Oct. 2017. https://arxiv.org/pdf/1707.03510.pdf

4) S. A. W. Shah, T. Khattab, M. Z. Shakir, M. O. Hasna, “Small cell association with networked flying platforms: novel algorithms and performance bounds,” IEEE JSAC SI UAV Communs., Submitted, Jan 1, 2018. https://arxiv.org/find/cs/1/au:+Shakir_M/0/1/0/all/0/1

NFPs for SON: Airborne SON

1) H. Ahmadi, K. Katzis, M. Z. Shakir, “A novel Airborne self-organising architecture for 5G+ networks,” in Proc. IEEE VTC-Fall’2017, Toronto, Canada, Sep. 2017. https://arxiv.org/pdf/1707.04669.pdf

2) E. Kalantari, M. Z. Shakir, H. Yanikomeroglu, and A. Yongacoglu, “Backhaul-aware robust 3D drone placement in 5G+ wireless networks,” in Proc. IEEE ICC FlexNet’2017, Paris, France, May 2017. https://arxiv.org/pdf/1702.08395.pdf

3) A. Omri, M. O. Hasna, M. Z. Shakir, “Mode Selection Schemes for D2D Enabled Aerial Networks,” Submitted, IEEE TVT, Feb. 2018. https://arxiv.org/pdf/1711.02220.pdf

4) A. Omri, M. Z. Shakir, M. O. Hasna, 3-D placement strategies of multiple UAVs in NFP-assisted terrestrial cellular networks, IEEE TWC, to be Submitted Feb. 2018. https://arxiv.org/find/cs/1/au:+Shakir_M/0/1/0/all/0/1

05/02/2018 27

05/02/2018 28

Thank You and Questions!

Muhammad Zeeshan Shakir Lecturer in Computer Networks / Programme Leader MSc Smart Networks School of Engineering and Computing University of the West of Scotland Room E368, E Block Building (South) | Paisley Campus, High Street | Paisley PA1 2BE, Scotland, UKTel: +44 (0)141 848 3420 Email: muhammad.shakir@uws.ac.uk