PFSUF: A Particle Filter Based Urban Radio Maps Localization Method for Sparse Users

Radio maps generated from crowdsourced data offer a cost-effective navigation solution in complex urban settings, by utilizing non-specialized sensors from user equipment (UE). However, urban radio maps can cause significant received signal strength (RSS) estimation errors and a single communication channel due to sparse UE, leading to inaccurate positioning. To address this challenge, we propose a particle filter method incorporating sparse users (PFSU) to exploit the typical radio maps based integrated positioning system. In the PFSU, a specialized kernel function is designed to dynamically estimate effective prior variance by leveraging radio maps and sparse UE positions. Thereby, a constraint function is developed to refine the prior distribution of the particle filter for enhancing positioning accuracy. Subsequently, the simulation experiments illustrate the effectiveness of the proposed method and its superior performance compared to other radio map-based localization algorithms reported in the literature.