Multi-agent DDPG Enpowered UAV Trajectory Optimization for Computation Task Offloading

In order to solve the problems of high cost, poor mobility and difficulty in coping with emergency in large-scale deployment of fixed edge computing nodes in mobile edge computing(MEC), an unmanned aerial vehicle(UAV)-assist task offloading algorithm is proposed to meet the need of computing-intensive and delay-sensitive mobile services. Considering constraints such as the flight range, flight speed of multiple UAVs and system fairness among users, the method aims to minimize the weighted sum of the average computing delay of users and the UAV's energy consumption. This non-convex and NP-hard problem is transformed into a partially observed Markov decision process, and we propose a multi-agent deep deterministic policy gradient algorithm to get optimal offloading decision and UAV flight trajectory. Simulation results show that the proposed algorithm outperforms the baseline algorithm in terms of fairness of mobile service terminals, average system delay and total energy consumption of multiple UAVs.