Research On Beamforming Techniques For Wireless Physical Layer Security With Imperfect CSI

Human society has been in the high efficiency information age owing to the rapid development of wireless communication techniques.The way of working and living has drastically changed.At the same time, the broadcast nature of wireless channels makes the transmitted messages vulnerable to eavesdropping. Hence, the security of the wireless communications is of prime importance. Physical layer security exploits the physical characteristics of wireless channels to enhance the security by signal processing and coding techniques. The perfect secrecy from the information theoretic perspective can be achieved. When the multi-antenna technique is adopted to create the advantage of the legitimate channel over the eavesdropping channel,the channel state information (CSI) at the transmitter plays a crucial role in determining the security performance. However, it is difficult for the transmitter to get the perfect CSI,due to many issues such as estimation or feedback errors.Hence, secure transmit design is a critical problem under different assumptions on channel models and the CSI, especially the imperfect CSI.The main achievements of this dissertation are listed as follows:1.We consider multiple-input single-output (MISO) wiretap channels in the presence of multiple single-antenna eavesdroppers.For the case where the channel covariance error matrices are bounded in elliptical uncertainty regions, we examine the worst case secrecy rate maximization (WCSRM) problem for one legitimate user scenario and the transmit power minimization problem (TPM) for multicasting. A robust artificial noise (AN) aided beamforming is proposed. The challenges of the above optimization lies in two aspects:1)there are infinite constraints owing to the elliptical uncertainty regions of error matrices;2) the optimization problem is nonconvex and can not be solved directly. To deal with the difficulties, we first transform the infinite constraints into equivalent positive semidefinite (PSD) constraints.Then the nonconvex problem is solved using the semidefinite relaxation (SDR) technique and a one-dimensional line search method.2.For the case where the errors are models as stochastic model,the TPM problem is studied under the broadcasting scenario.The reliability and the security are assured from the quality of service (QoS) perspective:the given probabilistic constraints of the signal-to-interference-plus-noise ratios(SINRs) at both legitimate users and the eavesdroppers should be satisfied. After reformulating the probabilistic constraints without explicit expressions into conservative deterministic constraints, the beamforming vector and the AN covariance matrix are obtained by solving a SDP.3.We consider multiuser MISO interference channels with confidential messages.Each transmitter delivers messages to the intended user, while keeping it secret from the remaining users. For the case with perfect CSI, the secrecy rate balancing problem is investigated, where the minimum secrecy rate is maximized.An iterative optimization based coordinated beamforming (CB) method is proposed.During each iterative process, a SDP is solved.The obtained solution is locally optimal.The convergence of the iterative algorithm is proved and checked through the simulations. Furthermore, the proposed frame is extended to the case with imperfect CSI, and a robust CB scheme is proposed.4. The multi-antenna relay channel in the presence of eavesdroppers is considered. When there is only one source-destination pair, the secrecy rate maximization problem (SRM) and the WCSRM problem are addressed for perfect CSI and imperfect CSI, respectively. The SRM problem is solved using the SDR technique and the two-level optimization algorithm, while the WCSRM problem is solved using the SDR technique and the successive convex approximation algorithm.5.When there are multiple source-destination pairs, we investigate the AN power maximization problem subject to the destination nodes’mean square error (MSE) constraints with imperfect CSI of destinations and without CSI of the multi-antenna eavesdropper. The beamforming matrix of the relay and the equalizers of destinations are jointly optimized through the iterative algorithm which is convergent.