Intelligent Reflective Surface Assisted Low-Bit Wireless Communication System Based on Genetic Algorithm

Intelligent Reflective Surfaces (IRS) have received great attentions in the field of wireless communications for optimizing the performance of systems. A majority of current researches assume that the IRS phase is continuous, which is not realistic due to hardware limitations. Those existing IRS discrete phase optimization methods are mainly based on ideal phase approximation by distance, which poses the problem of severe system performance loss at low precision quantization. This paper considers a multi-user downlink system assisted by an IRS with finite phase resolution. To find the optimal discrete IRS phase combination, we propose a genetic algorithm-based discrete phase shift optimization methodto improve the weighted sum-rate (WSR) of users in the case of low precision quantization. Simulation results show that the algorithm can effectively improve system performance compared to traditional optimization methods at low precision quantization.