Cooperative beamforming in multi-cells multi-users networks with quantized feedback

In this paper, we consider a downlink coordinated multi-cells multi-users network, in which each User Equipment (UE) knows its Channel State Information (CSI) perfectly and feeds back its quantized composite channel to the distributed Base stations (BSs) through a limited feedback. We propose a cooperative beamforming design, considering quantized composite channels, that reduces the inter-cell interference for both cases of perfect and outdated quantized CSI and an optimal power allocation that maximizes the system sum rate. Furthermore, we propose a bit labelling of codebooks, used for channel quantization, that outperforms the random labelling under channel feedback errors and delay. We analyze also the sum rate loss due to the outdated and quantized channel knowledge at the transmitter.