Trajectory Optimization and Computing Offloading Strategy in UAV-Assisted MEC System

Unmanned aerial vehicle (UAV) plays an important application scenario in mobile edge computing (MEC). In this paper, we jointly optimize the computing offloading strategy in MEC and the UAV trajectory to achieve the real-time communications between users and UAV, minimize the total energy consumption and delay, and improve users quality of service (QoS). As for computing offloading strategy, typically, UAV allocates channels to each user for data transmission. The energy consumption and delay generated by data transmission and computation are used as the evaluation criteria for offloading strategy. And the population diversity-binary particle swarm optimization (PDPSO) algorithm is used to obtain the optimal value. As for UAV trajectory optimization, the UAV communicates with the users in real time scenario. When the UAV arrives at a new position, it can receive the weighted sum of total energy consumption and delay based on computing offloading strategy. We aim to minimize the sum of energy consumption and delay at each position as the UAV completes a flight trajectory. Moreover, the deep deterministic policy gradient (DDPG) algorithm is used to obtain the optimal trajectory of the UAV. The simulation results show that our joint optimization algorithm is better than other joint algorithms.