A dynamic joint clustering scheduling algorithm for downlink CoMP systems with limited CSI

Abstract

In downlink coordinated multi-point (CoMP) systems, full cooperation among the base stations (BSs) in the network is unfeasible because of bandwidth and latency constraints in the backhaul and imperfect channel knowledge at the BSs. To deal with these issues BSs can be organized in clusters and perform CoMP transmission towards the user equipments (UEs) in their coverage area. In this work we develop a greedy dynamic joint clustering scheduling algorithm for the downlink of CoMP systems by assuming that the set of candidate clusters is limited and predetermined, based on UEs measurements. The set of non-overlapping selected clusters and the UEs scheduled in each cluster change dynamically at each time slot and are optimized by maximizing the system weighted sum rate. We evaluate the performance of the developed method by considering different criteria to select the set of available clusters and different policies to schedule the UEs in each cluster. Numerical results show the gain of the developed approach with respect to both single cell processing systems, i.e. when no coordination among the BSs is allowed, and CoMP systems with static clusters composed by the BSs belonging to the same site.