Combined User Selection and MIMO Weight Calculation for AP Cooperation in Dense Wireless Networks

Abstract

This paper addresses the problem of weighted sum rate (WSR) maximization in dense wireless networks with cooperative access points (APs) subject to a per-AP power constraint. We propose a combined optimization procedure that performs both user selection and MIMO weight calculation and scales well as the number of users increases. User selection eliminates some undesirable users, while MIMO weight calculation determines the precoders and combiners for all active nodes. A new performance metric, which takes into account available power, channel quality and orthogonality, and user weights, is used to perform an initial phase of user selection. A WSR maximization algorithm is then executed to optimize MIMO weights of selected users. The proposed algorithm includes additional user selection, i.e. certain users not eliminated in the first phase will be assigned zero-power stream during its execution. Numerical results show that our proposed algorithm achieves about 25% higher aggregate performance than the best existing algorithm while having a substantially lower running time. In fact, the running time is nearly constant as the number of users increases due to the very fast initial user selection phase.