Events

 
Seminar
Distributed Transceiver Design for the Multi-Antennas Interference Channel

by Dr. Zuleita Ka Ming Ho

Date
 :  25 Jan 2011 (Tue)
Time
 :  11:00 am - 12:00 noon
Venue  :  Rm 2463, 2/F (Lift 25, 26), Academic Complex, HKUST

Abstract

In this thesis, we aim to optimize transmit and receive strategies in a network where there is little or no centralized resource management unit and nodes have limited knowledge of the channel states and limited backhaul communication among each other. In particular, the transmitters have mostly locally available channel state information (CSI) and we assume no data sharing or joint transmit strategies design among transmitters, thus prohibiting joint MIMO transmission.

As transmitters must share the same system resources e.g. time and frequency, they generate interference in the process of communicating with the receivers making interference management essential. In a game theoretic perspective of the beamforming design problem over the interference channel, extreme egoistic and altruistic strategies can be defined. Egoistic transmitters act selfishly and maximize their own SINR despite the interference generated to the remaining of the network whereas altruistic transmitters exploit all resources to null out interference generated towards other receivers. It is intuitive to see that none of the above two strategies is generally sum rate optimal. A recent work shows that in the transmit beamforming vector design problem in the MISO-IC with single user decoding (SUD), balancing egoism and altruism brings the operating point on the Pareto boundary, which is the boundary of the achievable rate region assuming linear pre-processing. With this concept in mind, we investigate distributed transmit beamforming designs in the MISO-IC-SUD in Chapter 3.

As an extension of the beamforming problem on the MISO-IC-SUD, we design the transmit and receive strategies on the MIMO-IC-SUD in Chapter 4. Assuming mostly local CSIT, we model the problem as a Bayesian game which takes the unknown channel knowledge into consideration and where players maximize the expected utility function based on the statistics of the unknown channel. We derive the equilibria of the Bayesian games and revisit the sum rate maximization problem on the K users MIMO-IC-SUD. We observe that the sum rate maximization solution can be interpreted as a balance between the egoistic and altruistic equilibria. With this analysis, we provide an algorithm which allows the transmitters and receivers to optimize the transmit and receive beamformers iteratively.

Then we proceed to allow the receivers to have interference decoding capability (IDC) in the MISO-IC in Chapter 5. This additional degree of freedom allows receivers to decode interference and subtract it from the received signal and thus enjoy interference free communication when the channel realizations allow it. The intriguing problem is that the transmitters may design the beamforming vectors to amplify interference for easy interference decoding and removal which differs from the conventional design in interference mitigation systems.

Biography

Zuleita Ka Ming Ho was born in Hong Kong in 1984.  She had full scores in the Hong Kong Certificate Examination of Education in 2001 and enrolled into Hong Kong University of Science and Technology through the Early Admission Scheme for Outstanding Secondary School Students.  In 2003, she received the HSBC scholarship for Overseas Studies and visited Massachusetts Institute of Technology for 1 year.  In 2007, she received one of the biggest scholarships in Hong Kong, The Croucher Foundation Scholarship, which supports her doctorate education in EURECOM in Sophia Antipolis, France.  She received a Bachelor and a Master of Philosophy degree in Electronic engineering with First Class Honor at HKUST in 2004 and 2006. Under the supervision of Professor David Gesbert, she received with highest honor a doctorate degree with Telecom ParisTech in 2010.  Her research focuses on resource allocation problems with multi-cell cooperation on the physical layer: beamforming design problems on the interference channels, game theory.