Coordinated Multipoint Transmission in Dense Cellular Networks With User-Centric Adaptive Clustering

Abstract

Based on random network (RN) topologies generated from Poisson point processes (PPP), this paper investigates the performance of macrodiversity coordinated multipoint transmission (MD-CoMP) in dense cellular networks. First, the signal-to-interference-plus-noise ratio (SINR) outage probability is analyzed for a typical mobile station (MS) and for the global network. Next, a user-centric adaptive clustering method is described, which is designed to maximize each MS's normalized outage capacity (goodput). Simulation are carried out and show that MD-CoMP could significantly improve both the RN and regular hexagonal network (HN) coverage performance by increasing the tenth percentile of the SINR by 12 dB if each MS uses a CoMP cluster of size four. It is also shown that MD-CoMP is more beneficial for the RN since 78 MSs in the RN would choose CoMP to optimize their normalized goodput, whereas this number is 58 in the HN. Moreover, 58 MSs in the RN have their normalized goodput doubled compared with that with no CoMP, whereas this number is 36 in the HN. The impact of predefined clustering schemes is also evaluated, to show the importance of using a fully adaptive clustering to overcome cluster-edge issues, where the MSs' performance is poor due to the limited choices of BSs.