Dynamic TDD Interference Mitigation Using Graph Theory Based Cell Clustering in 5G Ultra-Dense Network

Abstract

Dynamic time-division duplex (TDD) is considered a promising solution to handle the unbalanced bursty and quick varied traffic in 5G ultra dense networks. However, dynamic TDD also bring out additional cross-link interference that may degrade the system performance. Since the network scenario of 5G becomes denser compared to LTE, the distance between user equipment (UEs) is smaller, UE-UE cross-link interference becomes more important and should be mitigated together with BS-BS cross-link interference. In this paper, we use cell clustering to deal with the cross-link interference. We find a better threshold for clustering and proposed a novel cell clustering algorithm based on graph theory. Simulation results show that the novel algorithm offers significant gain in UE SINR and UE rate compared with dynamic TDD without clustering and dynamic TDD with traditional clustering.