Performance analysis of MIMO cellular network with channel estimation errors

In this paper, we are interested in performance analyses of a multiple-input multiple-output (MIMO) multiple-access (MA) wireless network where Wyner's linear cellular array model is adopted. Notwithstanding some recent results have shown that, in the Wyner's model, the joint processing of the entire received signal with access point (AP) cooperation can dramatically enhance the system throughput because of the macrodiversity. However, in practice, the channel estimation quality from the providers of the macrodiversity is usually poor. Furthermore, existing literatures do not take channel estimation error into account. The channel estimation error provides us the motive to derive analytical solutions to the large-system throughput (or channel sum-rate) of the MIMO cellular network with imperfect channel estimations. Unlike the most large-system results, our derived results can not only be applied in the scenarios with any given input distributions (not necessarily Gaussian) from the mobile transmitters, but also clarify the performance of the network which was originally involved. Several issues we address including 1) how much of coherence interval should be spent for training and 2) the impact of AP cooperation is highlighted through the analytical results.