Research On Physical-layer Security In UAV Communications

With the technical developments,unmanned aerial vehicles(UAVs)are expected to play an important role in fifth generation(5G)cellular networks as special aerial devices/platforms to support ubiquitous communications,which significantly improve the rate,capacity,and coverage of conventional mobile communications networks.The widely deployed UAVs can be used not only as low-altitude mobile users that need to access to the mobile communication network,but also as aerial communication platforms to provide services for ground users.However,UAVs usually have dominant line-of-sight(Lo S)channels with ground nodes,which are more prone to eavesdropping attacks by malicious eavesdroppers,compared with traditional terrestrial communications.Recently,physical layer security is becoming a viable new solution to protect UAV communications against malicious eavesdropping attacks.Different from conventional cryptology-based security technology,physical layer security is able to provide perfect security for wireless communication systems from an information theory perspective.Therefore,it is emerging and of important practical and theoretical values to conduct research on physical layer security in UAV communications.In particular,this dissertation focuses on improving UAV communication security by using physical layer security under two different UAV communication setups.The main results of this dissertation are summarized as follows:1)First,the coordinated multipoint(Co MP)reception-enabled UAV communication system is considered.In this system,a joint optimization framework is proposed,which jointly optimizes the UAV's three-dimensional(3D)trajectory and transmission power allocation for maximizing the average security rate.However,the problem is non-convex problem and thus is difficult to solve.Towards this end,the alternating optimization technique is used to transform the problem into two sub-problems: trajectory and power allocation optimization.Then,the two sub-problems are solved by convex optimization and successive convex approximation(SCA),respectively.Finally,a local optimal solution of the original problem is obtained by iteratively updating them.The simulation verifies the effectiveness of the iterative convergence algorithm.Numerical results show that due to the spatial diversity gain employed in Co MP reception system,the UAV needs to optimize 3D trajectory to improve the security performance,which is different from conventional studies with only 2D trajectory optimization required.2)A dual secure UAV communication system with cooperative jamming is further considered.This system has a communication UAV and a nearby cooperative UAV.When the communicaion UAV delivers the confidential information to a ground node,the cooperative UAV cooperatively sends artificial noise(AN)to confuse the ground eavesdropper for protecting the confidentiality of the data transmission.The 2D trajectory and transmission power allocation of two UAVs are jointly designed to optimize the average security rate under the eavesdropper's location estimation error constraint.Because the objective function of the problem is an implicit and non-convex function,the problem can be converted into two sub-problems by using alternating optimization techniques after reasonable approximation processing.The two sub-problems are converted into two standard convex problems by the SCA,and a local optimal solution of the original problem is obtained after iteration.Numerical results verify the effectiveness of the algorithm and show that the average security rate is improved effectively by cooperative jamming,in which the communication and cooperative UAVs fly towards the ground node and eavesdropper for efficient communication and jamming,respectively.