Efficient Clustering for Multicast Device-to-Device Communications

Abstract

The purpose of this paper is to propose an efficient clustering method for multicast device-to-device (D2D) communications. By reviewing the related research works, we examine the physical and social characteristics that affect the clustering. Considering the quality of the radio link, the required data and files, and the number of neighboring users, we do the clustering and consequently find the cluster head and associated members. Using these criteria, a weighted metric is proposed which introduces the cluster head and assigns the members to each cluster. After clustering the users in each cluster, the cluster head multicasts the required data to the associated members. First, we propose an idea to improve the clustering performance, which offers a decrease in the number of clusters and an increase in the throughput. Then, two new clustering algorithms are proposed and they are compared in order to show the efficacy of the proposed ones. The numerical analysis using MATLAB shows that the proposed algorithms significantly reduce the number of required radio resources and sharply increase the throughput.