A Distributed Fault-Tolerant Mechanism For Mission-Oriented Unmanned Aerial Vehicle Swarms

Unmanned aerial vehicle (UAV) swarm consists of a number of autonomous and collaborative UAV nodes. The purpose of the research is to address the hypotheses: an efficient centralized resources management scheme should be adopted to meet the mission-oriented feature of UAV swarms, while a distributed fault-tolerant mechanism should be adopted to meet the characteristics of complex environment and address the issue of high failure rate of UAV nodes. By using an experimental modeling method, this paper proposes a management model and related mechanisms fitting for UAV swarms with the feature of "centralized decision making and distributed task execution." The model allows UAV nodes to manage their own resources autonomously and execute centralized decisions accordingly. The advantage of the model is that the master node of a UAV swarm is able to perceive all the resources of the system and the status of all (or some) of the UAVs efficiently; therefore, it provides a supportive environment for the UAV swarm application to make centralized decisions and optimize those decisions. Additionally, this paper proposes a distributed fault-tolerant model and related mechanisms to address the issue of high failure rate of UAV nodes, as well as the system vulnerability caused by harsh environment. Finally, this study developed a prototype system designed and implemented by OMNeT++ to confirm the truth of the hypotheses. The analysis of the experimental data and simulation results of the prototype validated the feasibility and usability of the distributed fault-tolerant mechanism and demonstrated its value for the mission-oriented UAV swarms.