KOUNTOURIS Marios

Thesis

Multiuser multi-antenna systems with limited feedback

The use of multiple antennas (MIMO) has been recognized as a key technology to significantly improve the spectral efficiency of next-generation wireless systems. In multiuser MIMO networks, the spatial degrees of freedom offered by multiple antennas can be advantageously exploited to enhance the system capacity, by scheduling multiple users simultaneously. However, these promising gains are highly sensitive to the amount of channel state information at the transmitter (CSIT).

In this thesis, we focus on linear beamforming and scheduling techniques relying on low-rate partial CSIT. We first propose to split the design between the scheduling and the final beam design stages. This two-stage approach is applied to a random beamforming (RBF) context and several refinement strategies, including beam power control, are proposed. Significant feedback reduction and throughput gains are observed, particularly in sparse networks.

In correlated MIMO channels, channel structure is exploited as a means to reduce feedback. We show how memory-based RBF can exploit channel redundancy and achieve close to optimum capacity for slow time-varying channels. In spatially correlated channels, statistical CSIT is combined with instantaneous low-rate channel quality information (CQI) and a maximum likelihood channel estimation framework is proposed.

Limited feedback strategies utilizing quantization codebooks are also investigated. The problem of CQI design is addressed and several scalar feedback metrics are proposed. It is shown that low-rate CQI feedback combined with channel directional information and greedy user selection can achieve a significant fraction of the optimum capacity.