Multi-UAV Cooperative Computational Latency Modeling and DDPG Optimization

Abstract

With the development of future mobile communication, how to provide better quality of service for latency-sensitive services by mobile edge computing (MEC) in unmanned aerial vehicle (UAV) is a hot issue. Therefore, this paper considers the cooperation of multi- UAV to establish a mobile edge computing network and proposes an optimized delay scheme under the cooperative computing of multi-UAV. In this network, two main works are done. The first work is to model the computational delay of the tasks after the block. The second work is to optimize the computational delay through the deep deterministic policy gradient (DDPG) algorithm. Finally, the simulation results showcase that the proposed scheme has high reliability from the reward function. When subtasks are transmitted and computed, optimal allocate bandwidth and computing resources can be obtained by minimizing the computing delay in the proposed scheme.