Physical Layer Security Performance Optimization of Unmanned Aerial Vehicle-Assisted Wireless Network

In the UAV-assisted wireless sensor network communication scenario, it is assumed that there is a UAV receiver, several ground sensors and a ground eavesdropper whose location is unknown. In order to improve the total security rate of the network, a physical layer security performance scheme based on TD3 algorithm is proposed to jointly optimize the trajectory and transmission power of the UAV. First, the approximate location of the eavesdrover is estimated based on Bayesian ranging and iterative least squares RSS positioning algorithm, and then the optimal and efficient UAV trajectory and sensor transmission power are selected to maximize the security rate. According to the derived safe rate expression, an improved double delay depth deterministic gradient strategy algorithm based on DDPG is proposed. The simulation results show that using TD3 algorithm to optimize UAV trajectory can effectively improve the safety speed of the system.

I. Introduction

Wireless Sensor network (WSN) [1] is widely used in different fields such as agricultural production, environmental monitoring, remote monitoring and intelligent transportation [2]. In recent years, WSN has been increasingly used to transmit private information [3]. Therefore, the realization of secure communication in WSN deserves attention. Physical layer security has become a complementary technology due to the large operational and storage overhead constraints imposed by OSI-based upper-layer solutions.