Multi-Cell Processing with Limited Cooperation: A Novel Framework to Timely Designs and Reduced Feedback with General Inputs

We investigate the optimal power allocation and optimal precoding for a multi-cell-processing (MCP) framework with limited cooperation. In particular, we consider two base stations (BSs) which maximize the achievable rate for two users connecting to each BS and sharing channel state information (CSI). We propose a two way channel estimation or prediction process. Such framework has promising outcomes in terms of feedback reduction and acheivable rates moving the system from one with unkown CSI at the transmitter to a system with instantanous CSI at both sides of the communication. We derive new extentions of the fundamental relation between the gradient of the mutual information and the MMSE for the conditional and non-conditional mutual information. Capitalizing on such relations, we provide the optimal power allocation and optimal precoding designs with respect to the estimated channel and MMSE. The designs introduced are optimal for multiple access (MAC) Gaussian coherent time-varying fading channels with general inputs and can be specialized to multiple input multiple output (MIMO) channels by decoding interference. The impact of interference on the capacity is quantified by the gradient of the mutual information with respect to the power, channel, and error covariance of the interferer. We provide two novel distributed MCP algorithms that provide the solutions for the optimal power allocation and optimal precoding for the UL and DL with a two way channel estimation to keep track of the channel variations over blocks of data transmission. Therefore, we provide a novel solution that allows with limited cooperation: a significant reduction in the CSI feedback from the receiver to the transmitter, and timely optimal designs of the precoding and power allocation.