Managing Interference in D2D Networks via Clustering and Topological Awareness

Abstract

In this paper, we address the application of topological interference management (TIM) on a clustered network of Device-to-Device (D2D) devices which are not aware of the surrounding channel state information (CSI), but only of the connectivity pattern. Our main objective is to develop a proper clustering algorithm that leads to increasing the total system degrees-of-freedom (DoF). For this, we model the interference network as a connected graph, transforming the clustering problem into a graph partitioning problem. To solve it, we base our method on the semidefinite problem (SDP) relaxation of the maximum-k-cut algorithm, while accounting for the maximum number of devices allowed inside the multiple-input multiple-output (MIMO) cluster environment. Simulation results show that proper clustering combined with TIM design renders TIM more scalable, and able to increase the system DoF.