Joint optimization trajectory and resource allocation for UAV-assisted semantic communications

Semantic communication can effectively save bandwidth, and enhance communication capabilities by transmitting semantic features. However, semantic communication has certain limitations, such as limited application scenarios and inflexible deployment. To this end, we investigate an unmanned aerial vehicle (UAV)-assisted semantic communication system in this paper. An UAV serves as a mobile base station to service users in designated area. Each user has different requirements for transmission delay and performance, and the UAV has a limited maximum flight time. We need to achieve the communication goals of all users in the shortest possible time, that is, to ensure that the information received by each user meets latency and quality requirements. This is a non-convex optimization problem, which is very complicated to solve using traditional methods. In order to solve this problem, we propose a deep reinforcement learning algorithm based on Proximal Policy Optimization 2. The simulation results confirm the effectiveness of our proposed algorithm.