DQN-Based Relay Deployment and Trajectory Planning in Consensus-Based Multi-UAVs Tracking Network

Currently, that multiple unmanned aerial vehicles (UAVs) complete the tracking tasks cooperatively has became an important development trend. However, in multi-UAVs tracking network, the mobility of tracking UAVs will seriously affect the performance of the network, resulting in frequent network interruption, prolonging convergence delay of consensus algorithm, and even unable to meet the tracking requirements. In this paper, we proposed an approach by deploying a relay UAV to improve the connectivity and consensus performance of the multi-UAVs tracking network. We divide the process of deploying relay UAV into two steps. Firstly, DQN-based deployment algorithm is proposed to obtain the initial position of relay UAV, whose energy consumption and consensus convergence delay of the tracking network are optimized. Secondly, during the movement of tracking UAVs, DQN-based trajectory planning algorithm is proposed, which effectively improves the consensus success probability and reduces the outage probability of the multi-UAVs tracking network. Simulations verify that the DQN-based algorithms have superior performances compared with existing works.