Analysis of cooperative MIMO networks with incomplete channel state information

Abstract

Coordinating the reception and transmission of signals across spatially distributed base stations has been shown to improve sum-rate performance by mitigating the effects of intercell interference in Multiple-Input-Multiple-Output (MIMO) cellular networks. Relying on recent results on the freeness of certain non-Gaussian random matrices, we provide an information theoretic analysis of cooperative MIMO networks. This analysis applies to the case where full channel state information is known at a subset of the bases and where statistical information is known at all others. Tools for evaluating random matrix transforms traditionally exploited in Mean Square Error (MSE) and mutual information analysis are provided, and the general model formulation paves the way for future work, where specific scheduling and/or power assignment schemes could be embodied in the newly presented framework.