A Link-Prediction Based Multi-CDSs Scheduling Mechanism for FANET Topology Maintenance

Abstract

In Flying Ad hoc Network (FANET), maintenance of topology is a quite difficult task due to rapid change of connectivity between flight nodes, i.e. unmanned aerial vehicles (UAVs). Aiming to this issue, in this article, we proposed a link prediction-based multiple Connected Dominating Sets (CDSs) scheduling mechanism for stable maintenance of FANET’s topology. In particular, a group of candidate CDSs are periodically scheduled for developing a stable backbone subnet of the topology. The proposed mechanism could achieve an early detection of topological changes by employing a Markov chain predicting model on the node’s mobility. The simulation results show that the proposed algorithm has a better success rate and less update overheads than the single CDS maintenance method, especially in a typical swarm pattern of UAV team.