Research On RSS-based Localization Algorithms Using Reconfigurable Intelligent Surface

Wireless localization has an increasing impact on auto driving,tracking,target supervision and other fields,and has been received widespread attention.According to the different target state,the existing algorithms can be separated into static target localization and dynamic target localization.At present,the mainstream static target location methods include multilateral localization method and subspace-based localization method,whose basic idea is to obtain location parameters through the distance,RSS,AOA and other location information.Using convex optimization,least square method and gradient descent optimization techniques to complete the precise localization of the target.The RSS is widely used because of its convenience and low complexity.Compared with static target localization,dynamic target localization is more difficult due to its own mobile characteristics,especially in the real-time requirement.This requires that the algorithm must be able to determine the location parameters of the target in a short time.At present,people have developed many efficient information integration technologies to address the moving target localization problem.The biggest characteristic of these technologies is that they can rapidly integrate multiple location information to achieve localization.Most of the existing static and dynamic target localization studies rely on far-field and line-of-sight conditions,but in actual scenes,targets may not be in the far-field region of the array,and line-of-sight links between targets may not exist.When the target is in the near-field region of the array and there is no line-of-sight link,the performance of the traditional localization algorithm may decline or even fail.Therefore,how to achieve accurate target localization in the near-field and line-of-sight link missing is an urgent problem to be solved.In the past two years,as a new technology,reconfigurable intelligent surface has attracted extensive attention in the field of wireless communication.Its outstanding advantage is that it can reflect incident signals independently by controlling elements on the surface.When the LOS links between nodes are severely blocked,the reconfigurable intelligent surface can create virtual line-of-sight links between nodes to bypass obstacles,thus effectively solving the problem of localization performance degradation caused by line-of-sight link loss.Based on the above advantages of the reconfigurable intelligent surface,combined with weighted least square method,alternate iteration and extended Kalman filter,this paper solves the problem of position parameter estimation in the case of line-of-sight link missing.The main contributions are as follows:（1）For the case that there is no line-of-sight link between anchor node and unknown node,this paper proposes a near-field RSS localization algorithm based on reconfigurable intelligent surface assistance.Specifically,a virtual line-of-sight link between the anchor node and the unknown node is constructed by using the feature that the reconfigurable intelligent surface can focus the signal in a predetermined direction.This feature addresses the problem that the performance of the localization algorithm decreases significantly due to the lack of line-of-sight link.In addition,this paper further discusses the problem of near-field localization in the scenario of line-of-sight link and virtual line-of-sight link coexistence,and proposes a reconfigurable intelligent surface parameter adjustment scheme to achieve high-precision localization.The simulation results show that the algorithms proposed in this paper can further improve the localization accuracy.（2）In view of the high complexity of existing localization algorithms,the proposed algorithms can obtain the distance and azimuth of unknown node simultaneously by using the maximum principle of RSS,which effectively reduces the dimension of search and reduces the computational complexity.The algorithms can simultaneously solve the problem of unknown node location and parameters optimization of reconfigurable intelligent surface by only azimuth search.On this basis,this paper deduces the FF or NF discriminant algorithm to reduce the adverse effects caused by the far-field assumption failure.The simulation results show that the near-far discriminant algorithm can further improve the localization accuracy.（3）Aiming at the problem of vehicle localization assisted by reconfigurable intelligent surface,this paper uses the vehicle position prediction results to adjust the phase of reconfigurable intelligent surface elements,so as to obtain the data containing position information,and effectively improve the vehicle localization accuracy in the absence of line-of-sight link.In order to solve the nonlinear problem of observation equation,this paper linearizes the nonlinear observation system through extended Kalman filter,completes the fusion of target position information,and realizes the dynamic localization of vehicles.Simulation results show the performance of the algorithm.