RC3: Robust Cognitive Cooperative Communications

Lead Research Organisation: University College London
Department Name: Electronic and Electrical Engineering

Abstract

According to the Ofcom's Digital Dividend report in 2007, spectrum is limited only because they are seriously underutilised due to rigid and inefficient management. It was reported that over 90% of locations could have around 100MHz spectrum available for other services. These underutilised or unused spectrum holes, also known as white spaces, are mainly due to the interleaved spectrum for the digital TV band. Given the fact that the entire 3G spectrum is only 75MHz, this is an unacceptable wastage. At the same time, however, this gives golden opportunities for mobile operators, broadband service providers and users in the UK to improve the QoE for personal communications using cognitive transmission in the spectrum TV white spaces.

In this context, BT is committed to exploiting the TV spectrum white spaces for providing wireless broadband access for homes in rural areas using cognitive radio (CR) technologies. According to BT, there are 2.75 million customers in rural areas, known as "not-spots" where, as ridiculous as it sounds, have no 3G coverage and Internet service is pretty much limited to dial-up access over residential or business telephone lines. In the "not-spots", the service is less than 2M bps but the TV spectrum white spaces, if utilised properly, can potentially cover more than 25% of the "not-spots" for improved services. The opportunity is that homes in a neighbourhood can share their antennas and signal processing capability to deliver much higher QoE using the spectrum white spaces at no extra cost.

This project takes a novel perspective of enhancing the energy and spectrum efficiencies of wireless communications via user cooperation (e.g., multiple homes cooperation), which offers the possibility to improve the channel by sharing the resources between users. This exceptionally challenging objective has the potential to redefine the architecture of wireless networks, provide a novel system solution for extending the coverage and enhancing the QoE of broadband communications.

In this project, the PI and BT (as the industrial partner) will join force to address the optimisation problem for cognitive cooperation. Our aim is to tackle the fundamental technical challenges specific to a cooperative MIMO channel. For instance, the required optimisation will need to take into account of individual users' requirements, constraints and fairness issues. Also, the proposed cooperative solution is also required to be robust to imperfect channel state information (CSI) and asynchronousity of the cooperating nodes, and be realised in a distributed manner. BT will be a key partner to provide invaluable inputs on the practical level to ensure that the project deliverables are exploitable. The final outcome of the proposed project will be the technologies for self-optimising cooperative antenna systems which can be used to provide broadband coverage for "not-spot" areas over wireless in the TV spectrum white spaces.

Planned Impact

The proposed research as a new innovative wireless communications solution based on user cooperation has demonstrable contribution to society and the economy. Impact of our research embraces diverse ways in which research-related knowledge and skills benefit individuals, organisations and nations [e.g., (1) researchers in wireless and broadband communications and other engineering problems dealing with uncertainties, (2) BT, Mobile VCE and other mobile service providers and (3) mobile service and broadband users in difficult environments such as rural areas, emergency situations, localisation and communications in battlefields, and high speed mobile video applications on board public transport, etc.] by fostering global economic performance, and specifically the economic competitiveness of the UK, increasing the effectiveness of public services and enhancing quality of life, health and creative output. In particular, impact of the proposed research can be categorised into:

- Knowledge: Techniques and scientific advances;
- People: Skills;
- Society: Quality of life;
- Economy: Inward investments, products and procedures.

In terms of knowledge, the proposed techniques have a new paradigm shift for providing wireless connectivity in harsh environments. As opposed to raising the transmission power and bandwidth, user cooperation is a whole new approach which is a flexible, yet effective, way to gain spatial diversity benefits that make communications not possible today possible. Also, it will open up a new dimension in designing communications systems where user fairness and distributed intelligence become essential and the scientific advances in this aspect will offer results that can go beyond wireless communications design. For instance, we see that distributed optimisation in the presence of measurement uncertainties which is the central challenge of user cooperation also appears in other engineering problems. The impact on scientific advances therefore will be tremendous.

Also, this project involves so much of curiosity-driven basic research elements as well as immediate industrial relevance. The project therefore will help the RA develop unique skills from understanding, designing to testing and implementing a wireless communication system. The experience and training the RA and the PI will get during this project will be essential for exploiting and commercialising research ideas.

In a longer term, user-cooperation technologies are believed to greatly enhance the quality of life resulting from improved mobile communications services. Because user cooperation is flexible and based on manipulating weaker signals, the gain in link performance will not compromise the network performance due to no increased interference. This makes it a particularly attractive solution for difficult environments such as emergency situations, disaster recovery, localisation and communications in battlefields, and high speed mobile video applications on board public transport, etc..

The industrial partner, BT, is keen to explore the feasibility of using cooperative communications in the spectrum TV white spaces for providing broadband services in rural areas in the UK. Please see the Impact Plan for the planned actions to maximise such impacts for the project.

Publications

10 25 50

publication icon
Hong Xing (2015) To Harvest and Jam: A Paradigm of Self-Sustaining Friendly Jammers for Secure AF Relaying in IEEE Transactions on Signal Processing

publication icon
Hu X (2018) Wireless Powered Cooperation-Assisted Mobile Edge Computing in IEEE Transactions on Wireless Communications

publication icon
Hu X (2019) UAV-Assisted Relaying and Edge Computing: Scheduling and Trajectory Optimization in IEEE Transactions on Wireless Communications

publication icon
Khandaker M (2015) Robust Secrecy Beamforming With Energy-Harvesting Eavesdroppers in IEEE Wireless Communications Letters

publication icon
Khandaker M (2015) Masked Beamforming in the Presence of Energy-Harvesting Eavesdroppers in IEEE Transactions on Information Forensics and Security

 
Description The project has led to the following key findings:

f1) Optimal transceiver design for simultaneous information and energy multicasting as well as broadcasting for a practically representative class of problem instances in both MISO (single- and multi-cellular) and MIMO systems. In general, it was unknown whether the beamforming design for problems with energy harvesting constraints would be optimal. However, in our works, we have been able to prove the existence of optimal solutions for many practically interesting problems.

f2) Optimal beamforming design for secrecy wireless information and power transfer considering both active and passive eavesdroppers. In particular, robust solutions are provided for scenarios where the eavesdroppers' CSI is only partially known.

f3) Incentive enabled user cooperation has also been proposed for MIMO relay systems where the relay nodes assist communication for guaranteed harvested energy in return.

f4) Full-duplexing schemes have been developed for simultaneous information and power transfer in which the detrimental self-interference turns to be a blessing. The residual self-interference (after cancellation) is, in fact, utilized to boost up energy harvesting.

For a practically representative class of problem instances in both single- and multi-cellular systems, one can obtain the exactly optimal resource allocation solution for SWIPT systems. Given the optimal solution, information and energy multicasting as well as broadcasting becomes more energy efficient towards the development of green communication systems.

A practical drawback of SWIPT is that the energy receivers (ERs) work at a higher power sensitivity level (e.g. -10 dBm) as opposed to the information receivers (IRs) (e.g. -60 dBm) which essentially means that the ERs need to have significantly stronger channels in order to harvest a useful level of energy. This generates additional information security threats particularly for systems where ERs are trusted only for harvesting energy, not for the private information destined to the IRs, since the ERs with better channels fading are more likely to be able to successfully decode the information. Our findings in f2) comes up with a practically appealing solution for this security issue with the aid of optimally designed artificial noise signals (AN). Note that for secrecy SWIPT systems, AN plays a doubly role namely i) keeping message secure to the IRs by jamming the ERs' reception, ii) properly designed AN beams can serve as energy beams to transport additional energy to the ERs so as to improve the amount of energy harvested.

In conventional cooperative networks, cooperative nodes may need to voluntarily sacrifice their valuable resources (e.g. power, frequency etc.) without getting anything in return. This may appear to be some sort of injustice to the volunteer nodes. However, with the advent of SWIPT, f3) actually developed schemes to provide the cooperative nodes at least some incentives in terms harvested energy for better motivation to cooperate. In that scheme, the helping nodes (relays) assist the communication only based on the harvested energy. Hence those volunteer nodes have nothing to lose if they cooperate during their idle period.

The SWIPT enabled full-duplexing (FD) schemes developed in f4) can bring enlightened future for FD systems. It is already known that with existing technologies the self-interference (SI) in FD systems cannot be fully cancelled, which remains as a major bottleneck of FD systems. However, in SWIPT enabled FD systems, at least part of the residual SI can be turned into blessings in terms of harvested energy. For battery-limited devices, this little amount of harvested energy can make a difference by prolonging the life-time of the communication.
Exploitation Route Our results are highly cited and taken up by peers as benchmarks.
Sectors Digital/Communication/Information Technologies (including Software)