Sub-optimal user scheduling for MU-MIMO system using channel quantization with two-stage feedback

We considered a multi-user MIMO broadcast (BC) system with multiple antennas at both transmitter and receiver sides. A scheduling scheme using semi-orthogonal user selection (SUS) with zero-forcing beamforming (ZFBF) is implemented. The total feedback load required in ZFBF systems is reduced by implementing a two stage feedback scheme by restricting the number of users to participate in each stage of scheduling. The minimal number of users participated in the scheduling process, the total channel state information (CSI) feedback from user side decreases drastically. Each user calculates its expected signal-to-interference-plus-noise ratio (SINR) and quantizes its channel direction information (CDI) vector. For this we use a codebook to generate the quantize CDI. A receiver combining strategy called the maximum expected SINR combiner (MESC) is used at the receiver side. In our approach we use the expectation of SINR rather than actual SINR of each user. The achieved sum-rate capacity of the proposed system is simulated and compared with existing ZFBF-SUS system. The effect of codebook size on the system performance is also investigated.