Joint Optimization Of Trajectory Planning And Secure Communication In UAV-assisted Air-to-ground Communication

Unmanned aerial vehicle（UAV）communication is a promising technology in 5G wireless communications.However,there are still some challenges in ensuring communication security such as the flight energy consumption of UAV and certain requirements for flight environment.In this paper,we consider a UAV-enabled communication scenario that a UAV needs to maintain secure communication with the ground communication nodes（GCNs）,subject to the known ground eavesdropping nodes（GENs）.To improve the performance of secrecy communication,the UAV can fly from the predetermined start point to the destination point.During the flight,it needs to select optimal communication positions,reduce the total energy consumption and avoid obstacles in reality environment.We formulate a UAV scheduling and path optimization problem（USPOP）based on the transmission model,the energy consumption model of UAV and the environmental constraints model.We solve this problem in two ways,one way is to decompose USPOP into two subproblems and propose an improved particle swarm optimization to solve multiple optimization objectives in the problem.The second way is to use the improved Non-dominated Sorting Genetic Algorithm Ⅲ（NSGA-Ⅲ）algorithm to solve this multi-objective problem.The specific operations are as follows:（1）We consider a practical secrecy communication scenario of UAV-enabled networks.Specially,the UAV flies from the start point to the destination point,and the UAV is regarded as a mobile BS to communicate with different GCNs,subject to GENs.It is noted that UAV needs to consider the environmental constraints.Then,we formulate a UAV scheduling and path optimization problem（USPOP）to jointly maximize the average secrecy rate and reduce the total energy consumption.（2）The part of USPOP is proven as NP-hard.Thus,USPOP is decomposed into two subproblems that are the UAV scheduling optimization problem（USOP）and the UAV path optimization problem（UPOP）.Then,a particle swarm optimization（PSO）with normal distribution initialization and differential mechanism（PSOND）and a PSOND with genetic mechanism and avoiding obstacles operator（PSONDGA）are proposed to solve the converted USOP and UPOP,respectively.（3）We consider USPOP as a multi-objective optimization problem,and the problem composed of three objectives: average communication secrecy rate,UAV hovering energy consumption and UAV flight energy consumption.Then,a NSGA-Ⅲ with discrete normal distribution initialization,differential mechanism,genetic mechanism and avoiding obstacles operator（NDGA-NSGA-Ⅲ）is proposed to solve USPOP.（4）The effectiveness and performance of the proposed PSOND,PSONDGA and NDGA-NSGA-Ⅲ are verified by simulation results.Additionally,the stability of the proposed algorithms performs better than other comparison algorithms.