Max-Min Fair Optimization in RSMA-Assisted Secure Communications With Artificial Noise
Huiyun Xia, Shuai Han, Cheng Li
Abstract
This work investigates the role of artificial noise (AN) in security constrained rate-splitting multiple access (RSMA) communication systems. By partially decoding the multi-user interference (MUI) and partially treat it as noise, the common stream in RSMA can serve as not only useful data for legitimate users, but jamming signals to confuse the external Eves. Considering this, whether introducing additional AN is beneficial to the RSMA-based communications remains unclear. To answer this question, we take user fairness into consideration and formulate a security constrained max-min fair optimization problem. The optimal transmit beamforming design and AN vector are then obtained by applying a joint line search and semidefinite relaxation approach based alternate optimization. Numerical results show that, facing eavesdroppers with strong capabilities, introducing AN helps improve the spectral efficiency of the system while satisfying the security constraints.