Convex Optimization Based Robust Spatial Multiplexing Techniques for Downlink Multiuser Wireless Systems

The demand for high speed data transmission over wireless access is growing steadily due to rich multimedia applications such as video streaming, music and interactive services. The expectation is to enable wireless access to provide the same data rate and quality of services (QoS) as that provided by wire-line counterparts. This requires provision of beyond 100 Mb/s and preferably 1Gb/s wireless access. The data rates of wireless systems can be increased by exploiting spatial diversity provided by multiple antennas at the transmitter and receivers. A basestation could serve multiple users simultaneously in the same frequency band using downlink spatial multiplexing techniques. For a frequency division duplex (FDD) based system, the use of multiple antennas at the transmitter however requires feedback of channel state information (CSI) from the receiver. Such channel state information when used at the transmitter will always be inaccurate due to channel estimation error, quantization of the estimates (due to finite budget for bit feedback) and relative motion between the transmitter and receiver. For example, when the channel is changing moderately fast, due to feedback delay, by the time the transmitter uses the channel state information, the true forward channel would have changed. The error which is the difference between the true channel and the estimates available at the transmitter could seriously degrade the transmitter diversity performance. Therefore, it is very important to consider the channel state information error when designing transmitter diversity techniques for the enhancement of capacity or coverage.Motivated by the rich theoretical and experimental results on the benefits of transmitter diversity techniques for wireless multiuser systems, we propose to develop advanced signal processing techniques to mitigate this very important and practical problem of imperfect channel state information at the basestation, a major limitation for using these diversity techniques in a highly hostile downlink channel environment. The research will be focused on developing robust beamformers/transmit diversity techniques that are resilient to channel state information error, using convex optimization techniques such as second order cone programming and semidefinite programming. The performance (in terms of bit error rate (BER), coded BER (CBER) and throughput) will be compared against conventional non-robust techniques for various channel fading profiles obtained using simulations and real field data provided through QinetiQ, facilitated by Professor Malcolm Macleod. This project has a distinct advantage of European collaboration with Prof. Alex Gershman who is a world renowned expert on robust beamforming and array processing techniques, and this window of opportunity should not be lost as we believe this provides an excellent vehicle for UK's standing in the field of robust design and array signal processing at an international level.