Research On Antenna Selection And Linear Precoding Algorithm In Secure Spatial Modulation System

Spatial modulation(SM)is emerging as a new and promising multiple-input multipleoutput(MIMO)communication technique,which exploits both the index of activated transmit and amplitude phase modulation signal to carry bit information.Compared to bell laboratories layered space-time architecture,the practical implementation complexity at the transmitter and receiver is significantly reduced,and the radio frequency cost is also reduced.Meanwhile,compared to space-time block coding architecture,the SM systems improve the spectral efficiency and energy efficiency.Thus,it is a promising MIMO technique and applicable to some future energy-efficient scenarios such as internet of things,wireless sensor networks and so on.However,it is very likely that the confidential messages are intercepted by unintended receivers,due to the broadcast nature and openness of wireless channels.Therefore,it is very important and necessary to address security issues for such SM systems.In this thesis,the key techniques for secure SM systems: transmit antenna selection and linear precoding methods,are investigated.The main research contents and contributions are organized as follows.(1)In order to achieve higher secrecy rate in secure SM-MIMO systems,two highperformance schemes of transmit antenna selection,maximizing signal-to-leakage and noise ratio(Max-SLNR)method and maximizing secrecy capacity(Max-SR)method,are proposed.Meanwhile,a conventional Euclidean distance-optimized antenna selection(EDAS)method is generalized to secure SM systems.By constructing the problem of Max-SLNR and solving it by using sorting-based method,the proposed Max-SLNR method significantly reduces the computational complexity.From simulation results,the two proposed methods achieve better performance of secrecy rate compared to the generalized EDAS method.Meanwhile,the secrecy rate performance of the proposed Max-SLNR method is close to that of the Max-SR method.More importantly,the Max-SLNR method requires a lower complexity,thus it achieves a better tradeoff between the performance of secrecy rate and complexity.(2)With secure SM-MIMO systems,the achievable secrecy rate does not have an easy-tocompute mathematical expression,and hence,which leads to high complexity in the optimal precoder design.To address this issue,an accurate and analytical approximation of secrecy rate is derived in this work.Using this approximation as the objection function,two low-complexity linear precoding methods are proposed,named as maximizing approximated secrecy rate gradient descend(Max-ASR-GD)and maximizing approximated secrecy rate successive convex approximation(Max-ASR-SCA).The Max-ASR-GD method converges to a local optimum by using gradient descend algorithm,and the Max-ASR-SCA method uses semidefinite relaxation and successive convex approximation techniques to deal with the nonconvexity in the precoder optimization problem and achieves near-optimal solutions.From simulation results and analysis,compared with the existing Maximizing secrecy rate gradient descend(Max-SR-GD)precoder design in the literature which directly uses the exact and numerically evaluated secrecy rate as the objective function,the two proposed designs have considerably reduced complexity.Meanwhile,the proposed Max-ASR-SCA design even achieves higher secrecy rate than the Max-SR-GD method.For example,the Max-ASR-SCA method even achieves a higher 0.6bps/Hz secrecy rate,in other words,an eighteen percent improvement compared to Max-SR-GD method,when signal-to-noise ratio is 5dB and the number of transmit antenna equals eight.