Joint Optimization of Resource Allocation and Trajectory in UAV-assisted Wireless Semantic communication System

Abstract

In this paper, we investigate the resource allocation design of an unmanned aerial vehicle (UAV)-enabled semantic communication system, in which UAVs are assigned to transmit semantic information to multiple user nodes. Our goal is to maximize the semantic information sum rate of the communication system by jointly optimizing the subcarrier allocation strategy and the trajectory of the UAVs, taking into account the minimum required semantic data rate for each user node, the minimum semantic similarity that can be received, the maximum cruise speed of the UAVs, and the initial/final position of the UAVs. The design is formulated as a generally tricky mixed-integer nonconvex optimization problem. Subsequently, a computationally efficient iterative algorithm is proposed to obtain a locally optimal solution, and further proofs demonstrate that the proposed algorithm is guaranteed to converge to a solution that satisfies at least the KKT condition of the original optimization problem.