Research On Physical Layer Secure Hybrid Precoding For Millimeter Wave Massive MIMO-SWIPT Systems

The millimeter-Wave(mm Wave)frequency band is the main frequency band for5G/B5 G to provide extended bandwidth for wireless transmission.Owing to its distinct advantages in line-of-sight(LOS)short distance transmissions,mm Wave has also become the ideal frequency band to implement the simultaneous wireless information and power transfer(SWIPT)technology.While the mm Wave multiple-input-multipleoutput(MIMO)SWIPT system can realize simultaneous transmission of information and power,the existence of the low security clearance energy harvesting receivers imposes new threats to information security of the communication system.Physical layer secure precoding technology can effectively improve the secrecy performance of the mm Wave MIMO-SWIPT system.It is therefore a promising research direction that can achieve endogenous security at the radio air interface.Because radio frequency devices operating at the mm Wave band are usually expensive,the commercial massive MIMO systems in the mm Wave band mostly adopt an analog/digital hybrid processing architecture.In this thesis,we investigate the physical layer secure hybrid precoding technology for mm Wave massive MIMOSWIPT system.The basic objective is to reduce the system's communication power consumption while ensuring balance/fairness of the energy users by designing the hybrid precoding scheme.The main contributions are summarized in the following:1)For the mm Wave multi-user multiple input single output(MISO)SWIPT physical layer secure communication system,in order to reduce the transmission power consumption,an optimization problem that minimizes the transmission power of the system by joint design of the digital and analog precoding subject to the constraints of confidential information transmission rate and harvested power was formulated.Since the optimization problem is a non-convex problem with coupled design variables,we decompose the problem into two sub-problems of digital precoding optimization and analog precoding optimization.By alternately solving the two sub-problems until convergence,solution to the overall problem can be obtained.Specifically,the digital precoding sub-problem can be solved by a two-layer optimization algorithm.The inner layer problem can be solved by the semidefinite relaxation(SDR)method,and the outer problem is solved by the golden section linear search method.The analog precoding optimization sub-problem,on the other hand,is readily solved by using a phase matching algorithm according to the characteristics of uniform modulus and random phase of the elements in the analog precoding.Simulation results show that the proposed hybrid precoding scheme has low complexity and good convergence property.It can effectively reduce the transmission power consumption of the system while satisfying the secrecy capacity and energy harvesting requirements.2)For the mm Wave multi-user MISO-SWIPT secure communication system,in order to improve the fairness in harvested energy among all users,a constrained maxmin optimization problem is formulated.The problem maximizes the minimum received power of the users' energy receiver by jointly optimizing the digital precoding,analog precoding,additive artificial noise(AN)and power allocation subject to the constraints of confidential information transmission rate and the total transmission power.The optimization problem is shown to be a non-convex problem with the design variables coupled in both the objective function and the constraints.We decompose the problem into two sub-problems of analog precoding optimization and digital precoding optimization,then solve them separately.Specifically,the analog precoding subproblem can be solved by a codebook-search method,while the digital precoding subproblem can be transformed into a linear matrix inequality(LMI)and second-order cone(SOC)form through relaxation,Taylor expansion and second-order cone transformation techniques.An iterative successive convex approximation algorithm is then developed to find suboptimal solutions to the problem.Effectiveness and convergence of the proposed algorithm are verified through extensive simulation with different systems setups.