Study On Physical Layer Security In UAV Communication

The high maneuverability of the unmanned aerial vehicle(UAV),facilitating fast and flexible deployment of communication infrastructures,brings new and valuable opportu-nities to the future wireless communication industry.Nevertheless,UAV communication networks are faced with severe security challenges since air-to-ground(A2G)communica-tions are more prone to eavesdropping attacks than terrestrial communications.In contrast to conventional security methods,which necessitate secret keys whose management and distribution may result in security vulnerability,physical layer security guarantees trans-mission secrecy by exploiting the channel characteristics.The dominant line-of-sight links in the A2 G channels and the mobility of UAVs,being deciding factors of the channel qual-ity,give rise to not only opportunities but also challenges from the perspective of physical layer security.Physical layer security in UAV communication is just in its infancy stage and lots of relevant problems remain to be investigated.This dissertation first investigates the average secrecy rate maximization(ASRM)problem.Since the mobility of the UAV brings new degrees of freedom for network de-sign,this dissertation aims to optimize the transmission power and the UAV trajectory for ASRM.Specifically,the considered system includes a source,a destination,an eaves-dropper,and a UAV-enabled mobile relay.The resulting ASRM problem is a large-scale nonconvex problem.To deal with it,this dissertation studies its power allocation subprob-lem and trajectory optimization subproblem.As for the power allocation subproblem,this dissertation proposes a concave-convex procedure(CCCP)-based iterative algorithm,and derives a semi-closed solution for the CCCP iteration.When the UAV trajectory is given as a special symmetric trajectory,this dissertation proves that the information-causality constraint can be removed without loss of optimality,and obtains the optimal transmis-sion power at the source in closed form.As for the trajectory optimization subproblem,this dissertation proposes a sequential convex programming(SCP)-based algorithm.For a special scenario where the relay's transmission capability is the system bottleneck,this dissertation proves that the trajectory optimization subproblem becomes a UAV place-ment problem,and obtains the optimal placement location in closed form.According to the power allocation algorithm and the trajectory optimization algorithm,this dissertation proposes an alternating optimization(AO)algorithm for ASRM.Numerical results verify that the proposed algorithm can effectively enhance the average secrecy rate.Then,this dissertation studies the UAV's secure transmission from the aspect of se-crecy energy efficiency,which is an important performance criterion in physical layer security.Since the UAV propulsion energy is huge and the on-board energy is limited,secrecy energy efficiency is even more important for physical layer security in UAV com-munication.This dissertation optimizes the UAV trajectory and transmission power for secrecy energy efficiency maximization(SEEM).The considered scenario is consisted of a fixed-wing UAV,a destination,and an eavesdropper.The resulting SEEM problem has a complex structure.To tackle it,this dissertation considers its power allocation subprob-lem and trajectory optimization subproblem.As for the power allocation subproblem,this dissertation proposes a CCCP-based iterative algorithm,and gets a closed-form solution for the CCCP iteration.As for the trajectory optimization subproblem,this dissertation proposes an algorithm according to SCP and Dinkelbach's method.Based on the power allocation algorithm and the trajectory optimization algorithm,this dissertation proposes an AO algorithm for SEEM.Numerical results verify that the proposed algorithm can effectively improve the secrecy energy efficiency.Last,this dissertation investigates the UAV's secure transmission from the aspect of secure multicasting.In traditional multicasting system,the performance is limited by the bottleneck link between the transmitter and its furthest user.UAV-enabled multicasting can overcome this problem through adaptive trajectory design.In the considered scenario,a UAV is dispatched to transmit common information to multiple legitimate users(LUs)with the presence of multiple eavesdroppers.Due to the limitation of the bottleneck link,it is difficult for a stationary transmitter to complete the secure multicasting task.This dis-sertation proposes a virtual base stations(VBSs)-based secure transmission scheme.The UAV successively flies to different locations to act as multiple virtual base stations,each of which serves a group of LUs.According to the proposed scheme,this dissertation es-tablishes an advanced N-center(ANC)problem,which aims to partition LUs into groups and optimize the placement of the VBSs.The ANC problem is NP-hard.In order to deal with it,this dissertation proposes a stack-based depth first branch-and-bound algorithm to search the possible user partitions.For a given user group,an AMEB algorithm is pro-posed to optimize the placement of the corresponding VBS.Next,the traveling salesman problem(TSP)algorithm is employed to optimize the visiting order of the VBSs.Numer-ical results verify that the proposed algorithm can effectively reduce the time required to complete the secure multicasting task.In order to deal with the eavesdropping threats in UAV communication,this dis-sertation makes an intensive study of physical layer security.By utilizing the convex optimization theorem,the branch-and-bound algorithm,and the TSP algorithm,this dis-sertation provides solutions for secure transmission in UAV communication from the as-pects of ASRM,SEEM,and secure multicasting,facilitating the wide application of UAV communication systems.