Physical Layer Security Analysis Based On Equivalents Theory And Spectral Analysis Theory Of Random Matrix

Physical layer security(PLS),which is an important supplement to the traditional encryption method,can achieve "perfect secrecy" for wireless com-munication and shows great value in the aspect of enhancing the secrecy per-formance of wireless communication systems.At present,the PLS technol-ogy of the single eavesdropping model has been studied extensively.However,wireless communication systems may be threatened by hybrid eavesdropping,e.g.,the coexistence of passive and active eavesdropping,due to its open na-ture.The results derived by single eavesdropping model cannot evaluate this scenario sufficiently.This dissertation analyzes the physical layer secrecy per-formance of different hybrid eavesdropping models based on the deterministic equivalents theory(DET)of random matrix in wireless communication scenar-ios with ultra-dense deployment and large-scale antenna features.Meanwhile,for the active eavesdropping user detection in wireless communication systems,a PLS detection method is proposed on the basis of the spectral analysis theory(SAT)of random matrix.The main contributions and innovations are summarized as follows:(1)In the single-cell multi-user wireless communication scenarios with passive and active eavesdropping,the deterministic equivalent results of the signal-to-interference-plus-noise ratios(SINRs)of legitimate and eavesdrop-ping users are derived based on the DET of random matrix.Then,the impact of parameters has be analyzed such as the relationship between the number of base station antennas and the SINRs.Furthermore,the secrecy outage probability(SOP)and average secrecy rate(ASR)of half-duplex eavesdropping scenario are derived,and those of full-duplex eavesdropping scenario are derived simi-larly.The Chebyshev-Gauss quadrature method is used to obtain approximate closed-form expressions of SOP and ASR.(2)In the single-cell multi-user wireless communication scenarios with internal and external eavesdropping,a technical scheme of the regularized chan-nel inversion(RCI)precoding under the cover of artificial noise is proposed to enhance the secrecy performance.Based on the DET of random matrix,the closed-form expression of SOP is derived,which reveals the relationship be-tween the SOP and the number of base station antennas.The numerical and simulation results show that the ASR can be improved by 1.7 times with appro-priate artificial noise power.(3)In the multi-cell and multi-user ultra-dense wireless communication scenario,an artificial noise based secrecy transmission scheme is proposed.Based on the DET of random matrix,the closed-form expression of the sys-tem ergodic secrecy rate is derived.The convex optimization theory is used to derive the optimal artificial noise power allocation factor.The numerical re-sults show that the secrecy performance can be greatly improved while artificial noise technology and the optimal power allocation factor are applied.(4)For the active eavesdropping user detection in wireless communica-tion systems,a PLS detection method is proposed on the basis of the SAT of random matrix.The effectiveness of the proposed method has been verified by simulation.