RIS-Aided Indoor Positioning System Based on Passive Reflective Elements Optimization

With the development of 6G technology, Reconfigurable Intelligent Surface (RIS), as one of the core technologies, has received wide attention in recent years. It not only improves the performance of the wireless communication system but also constructs a new localization scenario that utilizes the RIS as assistance. The RIS-aided positioning system can improve the localization performance of the Non-Line-of-Sight (NLoS) environment by building Virtual-Line-of-Sight (VLoS) links. However, it also faces challenges. First, the Cramér-Rao Lower Bound (CRLB), which is usually utilized to evaluate the system performance, has been analyzed only based on one or two parameters in existing research. It is not precise enough under the practical situation. Second, since there are many passive reflective elements on the RIS, applying all of them to localization will not only increase the storage space but even decrease the localization accuracy due that some elements may provide ambiguous information. Therefore, in order to solve the above problems, we first analyzed the CRLB of multi-parameter estimation to be consistent with the real environment. Second, a cutting-edge metaheuristic optimization algorithm, i.e., Artificial Rabbits Optimization (ARO) algorithm, is proposed to optimize the passive reflective elements, which aims to minimize the CRLB to increase the location accuracy while reducing the storage simultaneously. The experimental results show that the multi-parameter CRLB analysis is more in line with the actual situation. In addition, after the RIS passive reflective elements optimization, it improves the location accuracy by about 15.26$\%$ compared with using all elements.