Sum rate analysis and quantizer design for a quantized heterogeneous feedback MIMO OFDMA downlink

In this paper, we analyze the effect of quantization on a multiuser MIMO OFDMA downlink employing heterogeneous partial feedback design method, which adapts users´ feedback resources to their frequency domain channel statistics. This adaptive feedback design is investigated for the best-M partial feedback scheme, and the closed form expression for the sum rate under finite-rate quantization is derived for a system employing transmit antenna selection/maximum ratio combining (TAS/MRC) spatial diversity scheme under the generalized Nakagami fading channels. The sum rate provides a basis for determining the different quantization levels of an optimal non-uniform quantizer. For complexity reasons, a low complexity near optimal quantization strategy is proposed which employs uniform quantization with optimized intervals.