Research On Secure Beamforming And Cooperative Jamming For Correlated Main And Wiretap Channels

Due to the broadcast nature,security has been always a critical issue for wireless communications.By exploiting the random nature of physical layer transmission media,physical layer security(PLS)techniques,through proper coding and signal processing,ensure that confidential messages can be decoded with lower complexity only by the destination.With the rapid development of 5G and the Air-Space-Ground Communication Networks,it is indisputable that the correlation between the main and wiretap channels occurs.Since the correlation narrows the difference between the received signal quality at the destination and the eavesdropper,there is no doubt that the secrecy performance must be degraded.Currently,artificial noise(AN)-aided beamforming and cooperative jamming are mostly based on the premise that the main and wiretap channels are mutually independent.If such a restriction is liberalized to the case of correlated main and wiretap channels,the resulting secrecy performance is undoubtedly impaired.Concerning these issues,AN-aided beamforming and cooperative jamming for correlated main and wiretap channels are investigated in two cases of perfect legitimate channel state information(LCSI)and imperfect LCSI,aiming at reducing the secrecy loss due to the correlation between the main and wiretap channels to achieve a better secrecy performance and mitigate the security threat.To enhance the secrecy for correlated main and wiretap channels,it is necessary to establish the model of correlated main and wiretap channels and analyze its performance theoretically,which is the premise of the research on beamforming and cooperative jamming.Firstly,the relationship between the main and wiretap subchannels is deduced based on the joint power probability density function of the correlated subchannels.Based on this,simple and practical formulas precisely describing the multiple-antenna channel model of correlated main and wiretap channels are derived,including correlated MISO wiretap channels with perfect LCSI and imperfect LCSI,and cooperative-jamming-aided correlated MISO wiretap channels with perfect LCSI and imperfect LCSI.Next,the cumulative probability density function of the signal-to-interference-plus-noise-ratio(SINR)at the destination and eavesdropper are approximately given in a form of series or in a closed form.For correlated multiple-input-single-output(MISO)wiretap channels with perfect LCSI,a correlation-based AN-aided beamformer is designed so as to maximize the achievable secrecy rate under secrecy outage constraint.Firstly,the maximization of secrecy rate under secrecy outage constraint is considered for correlated MISO wiretap channels with perfect LCSI.Because the closed-form expression for the probabilistic constraint is not available in the nonconvex optimization problem,two typical approaches,namely the worst-case bounding approach and the Bernstein-type inequality approach,are employed to conservatively and approximately transform the probabilistic constraint into a deterministic form,based on which a convex optimization problem is formed and a safe approximate solution of beamforming vector and AN covariance matrix can be obtained.Next,the achievable secrecy rate under secrecy outage constraint is approximately given in a form of series.In addition,the study is also extended to a specific scenario of multiple-antenna eavesdroppers.For cooperative-jamming-aided correlated MISO wiretap channels with perfect LCSI,a correlation-based cooperative-jamming-aided beamformer is designed so as to maximize the achievable secrecy rate under secrecy outage constraint.Firstly,the maximization of secrecy rate under secrecy outage constraint is considered for cooperativejamming-aided correlated MISO channels with perfect LCSI.To make the optimization problem feasible,the worst-case bounding approach and the Bernstein-type inequality approach are employed to conservatively and approximately transform the probabilistic constraint into a deterministic form,based on which a safe approximate solution of beamforming vector and AN covariance matrix can be obtained.Next,the achievable secrecy rate under secrecy outage constraint is approximately given in a form of series.In addition,the study is also extended to a specific scenario of multiple-antenna eavesdroppers.For correlated MISO wiretap channels with imperfect LCSI,a correlation-based robust AN-aided beamformer is designed under the premise of SINR-outage-probability(SINR-OP)assurance.Firstly,the research on beamforming for correlated MISO wiretap channels with perfect LCSI is extended to the case of imperfect LCSI.As the correlation between the main and wiretap channels damages secrecy,the achievable secrecy rate under high correlation is small not to satisfy some high-rate applications with less rigorous secrecy required.Thus,by exploiting the correlation and employing a novel criterion of SINR-OP,three optimization problems are formulated.By employing the worst-case bounding approach and the Bernstein-type inequality approach to conservatively and ap-proximately transform the probabilistic constraint into a deterministic form,a safe approximate solution of beamforming vector and AN covariance matrix can be obtained.For cooperative-jamming-aided correlated MISO wiretap channels with imperfect LCSI,a correlation-based robust cooperative-jamming-aided beamformer under the premise of SINR-OP assurance is designed.Firstly,the research on cooperative jamming for correlated MISO channels with perfect LCSI is extended to the case of imperfect LCSI.Then,by using the criterion of SINR-OP,three optimization problems are formulated.By employing the worst-case bounding approach and the Bernstein-type inequality approach to conservatively and approximately transform the probabilistic constraint into a deterministic form,a safe approximate solution of beamforming vector and AN covariance matrix can be obtained.The results in this work show that the secrecy performance for correlated main and wiretap channels can be improved by the correlation-based AN-aided beamforming and cooperative jamming.