Secure Rate Maximization for Reconfigurable Intelligent Surface Assisted Transmission with Low-resolution DACs

This paper investigates the secrecy rate of multiple reconfigurable intelligent surfaces (RISs)-assisted millimeter wave system with low-resolution digital-to-analog converters, where a base station (BS) serves a single-antenna user while a single-antenna eavesdropper wiretapping the information sent to the user. To reduce energy consumption, the RIS that can provide maximal secrecy rate is selected to assist transmission. With the consideration of the quantization noise, we design the transmit beamforming at the BS and RIS phase shift jointly to maximize the achievable secure rate subject to transmit power constraint and unit-modulus constraint. However, it is a non-convex fractional programming problem. To tackle this non-convex problem, the alternative optimization algorithm is introduced to decouple the variables. Specifically, the quadratic transform technique is adopted to design the transmit beamforming, and the phase shift of RIS is obtained by generalized Rayleigh quotient algorithm. Numerical results demonstrate that considerable secrecy rate can be achieved without significant loss caused by quantization noise in the proposed algorithm.