Secure Transmission in Reconfigurable Intelligent Surface Aided Cooperative Cognitive Radio System

Reconfigurable intelligent surface (RIS) draws great attention due to its unique nature of changing the uncontrollable electromagnetic environment, and it has been verified in the existing study that in cognitive radio (CR) system, without excessive interference to primary user (PU), RIS greatly helps enhancing the secure transmission performance of secondary user (SU). However, existing studies do not consider the security issue of both PU and SU in RIS assisted CR system, which remains a question worth studying. Hence, this paper considers a cooperative CR system, and we aim at jointly enhancing the transmission performance of both PU and SU under eavesdropping attacks on them. Specifically, a non-convex sum-rate maximization problem of PU and SU subject to secrecy outage probability constraints under imperfect channel state information (CSI) with statistical estimation error between RIS and eavesdropper is formulated. And an efficient successive approximation in combination of alternating optimization algorithm is proposed to tackle this problem so that a local optimal solution of beamformer at PU, SU as well as phase shift coefficient at RIS can be obtained. Numerical examples demonstrate the significant enhancement of performance on our algorithm compared with the benchmark scheme of random phase shift coefficient at RIS.