Weight based dominating set clustering algorithm for small satellite networks

Abstract

Due to the size and density of small satellite networks (SSNs), the traditional clustering algorithms of large monolithic satellite networks are often limited. This paper proposes a novel distributed weight-based dominating set clustering algorithm to address the clustering problems in the stochastically deployed SSNs. Considering the unique features of small satellites, this algorithm is able to form the clusters efficiently and stably. In this algorithm, satellites are separated into different groups according to their spatial characteristics. Firstly, a minimum dominating set is chosen as the candidate cluster head set based on their weight, which is a weighted combination of residual energy and connection degree. Then the cluster heads admit new neighbors that accept their invitations into the cluster, until the maximum cluster size is reached. Evaluated by the simulation results, in a SSN with 200 to 800 nodes, the algorithm is able to efficiently cluster more than 90% of nodes in 3 seconds.