Load-Aware Greedy Dynamic CoMP Clustering Mechanism for DPS CoMP in 5G Networks

Abstract

Cooperative communications offered by the Coordinated Multipoint (CoMP) technique have been introduced and become an important technique in the 5G mobile networks to support very high system capacity and improve the end-user service quality, especially for the cell-edge users via the dynamic coordination of transceivers at different geographical sites. One major issue driving the performance of CoMP operation is the decision on the coopering clusters or the cluster formation strategies. In this work, an approach of CoMP clustering has been studied and proposed, the so-called load-aware greedy dynamic CoMP clustering mechanism. Our proposed CoMP clustering technique is deployed on top of the actual traffic-based load-aware Dynamic Point Selection CoMP (DPS CoMP) mechanism, which is our previously proposed mechanism. The approach is aimed at maximizing spectral efficiency with appropriate cell coordination based on well-balancing cell loads within the coordinating area of the 5G networks. System performance is numerically studied and observed for the 5G Non-Orthogonal Multiple Access (NOMA) systems embedded with our proposed mechanisms for the realistic scenarios, i.e., homogeneous network and Heterogeneous Network (HetNet) cases. Numerical results illustrate the benefit brought by our approaches in comparison with others CoMP clustering techniques as well as the potential for further improvement.