Research On The Design Of Intelligent Reflecting Surface Assisted Secure Beamforming With Partial CSI

With the ongoing improvement of users’ communication requirements and the rapid growth of the number of users,while pursuing high-speed and low-power communication,the demand for communication security is also gradually increasing.Therefore,how to ensure the security of communication transmission in high-speed and low-power system is of vital interest to individuals and even countries.As important security techniques,physical layer security techniques take advantage of the special properties of wireless channels and do not rely on the computational resources and network parameters of communication devices.They have the advantages of low complexity and resource saving.However,in the case where the useful channel is strongly correlated with the eavesdropping channel,the existing physical layer security techniques cannot guarantee secure transmission.In order to overcome the above difficulties,intelligent reflecting surface(IRS)can be effectively combined with physical layer security techniques.Taking advantage of the special structure,green energy saving and low complexity of intelligent reflecting surface,the wireless transmission channel can be reconstructed by changing the phase shift and amplitude information of the reflected signal from reflection units to avoid the strong correlation of the channel,so as to realize wireless secure transmission.Since intelligent reflecting surface does not have the function of signal processing,it is hard to acquire perfect channel state information(CSI).Based on this,this paper studies the design of intelligent reflecting surface assisted secure beamforming with partial channel state information.The principal research and work of the paper are shown below:First of all,for the situation where there is an eavesdropper in the link between the transmitter and the legitimate user,this paper establishes an IRS-assisted single-user singleeavesdropper wireless communication system model.Under the condition that the transmitter can only acquire partial CSI,with the ultimate goal of maximizing the secrecy rate of the system,the beamforming vector at the transmitter and the IRS phase shift matrix are jointly optimized.For the optimization problem,this paper uses the feature of reconstructed channel de-strong correlation to transform the original non-convex problem into two subproblems: IRS passive beamforming subproblem and transmitter active beamforming subproblem.Aiming at these two subproblems,this paper proposes the Alternating Direction Method of Multipliers(ADMM)and the Weighted Minimum Mean Square Error(WMMSE)algorithm respectively to solve them.Simulation results demonstrate that the proposed approach can effectively enhance the security of the system.Secondly,considering the scenario of multiple eavesdroppers around a legitimate user,an IRS-assisted single-user multi-eavesdropper wireless communication system model is constructed.Under the condition that partial CSI is known at the transmitter,an optimization problem of maximizing system secrecy sum rate is established,and the IRS phase shift matrix and transmitter beamforming vector are optimized.Since the original optimization problem is a non-convex problem with constant modulus constraints,in this paper,the original non-convex problem is transformed into two subproblems to be solved sequentially by using the de-correlation property of the reconstructed channel.The two subproblems are the IRS passive beamforming subproblem and the transmitter active beamforming subproblem,respectively.For the IRS passive beamforming subproblem,ADMM algorithm is used to solve it.For the transmitter active beamforming subproblem,two optimization algorithms are proposed: 1)WMMSE algorithm that achieves good performance;2)Equivalent Channel Decomposition(ECD)algorithm that provides comparable performance to WMMSE algorithm at much less complexity.The final simulation results indicate that proposed method has excellent convergence characteristics,and meanwhile can markedly improve the security sum rate and security performance of the system.