On the Optimal Discrete Phase Values of Reconfigurable Intelligent Surfaces

Design of reconfigurable intelligent surfaces (RISs) should find the optimal phase shift matrix and the optimal phase shift values. By encoding the RIS phase shifts and the phase shift values into a design matrix, a new RIS design scheme with multiobjective optimization using MOEA/D-RIS is proposed. The scheme can search for the optimal RIS phase matrix and the optimal phase shift values in the meantime. For multi-user RISassisted MISO system, binary phase shifts of values of 5°and 166°, other than 0°and 180°, will yield better RIS performance.