Reconfigurable Intelligent Surface-aided Security Precoding Scheme Based On Symbolic Information

With the development of wireless communication technology,the communication needs of high speed,high capacity,high efficiency,and high security is also growing.Due to the inherent broadcast nature of wireless signals,there are always various security challenges in communication.In recent years,reconfigurable intelligent surface(RIS)has been widely used to enhance physical layer security in wireless communication systems.In addition,symbol-level precoding(SLP)has recently attracted much attention because of its great advantage in exploiting multi-user interference.Therefore,with the rapid development of communication technology and the decreasing cost of hardware,the combination of RIS and SLP is not only crucial for research in the field of physical layer security,but also offers a better solution to the problem of insecurity arising from the ever-increasing intelligence of the eavesdropper.This thesis focuses on the SLP problem of RIS-assisted secure communication,including secure pre-coding scheme in the presence of the common eavesdropper or the intelligent eavesdropper.The main work is as follows.For the common eavesdropper,i.e.,the eavesdropper,intuitively decodes the message signal without any manipulation,this thesis derives the transmit power minimization problem by introducing a joint design of SLP and RIS,using the widely used assumption that RIS satisfies the maximum amplitude and phase continuity in order to satisfy the RIS reflection coefficient constraint,while modelling and analysing the symbol error rate(SER)of the user and the common eavesdropper.For the non-convex of the problem,the thesis proposes an alternating optimization algorithm based on alternating direction method of multipliers(ADMM)and constraint projection.The simulation results verify that the proposed algorithm is effective in reducing system power in multiple input single output(MSIO)scenario and demonstrate that the proposed algorithm can further improve system security.For the intelligent eavesdropper,i.e.,eavesdropping by means of a transmission scheme that mimics BS,this thesis introduces Artificial Noise(AN)and considers joint SLP and AN-based RIS design.As it was found that the solution provided by minimizing power may not be optimal in terms of security,this thesis reconstructs the minimizing smart eavesdropper SINR problem.For the nonconvex of the problem and the large number of constraints and variables,the thesis proposes the block coordinate descent(BCD)algorithm.We transform it into an approximate problem,which is solved by the semi-positive definite relaxation and the Gaussian randomisation method.The simulation results verify that in the presence of an intelligent eavesdropper,the proposed algorithm can further suppress the eavesdropping capability of the intelligent eavesdropper compared to the conventional algorithm without RIS assistance,regardless of whether the full channel state information(CSI)or statistical CSI of the eavesdropper is known to BS,thus ensuring the users’ communication security.The research results are of great significance to enrich enriching existing physical layer security methods and promote low-energy secure wireless communications.