An Indoor Rfid Location Algorithm Based On Support Vector Regression And Particle Swarm Optimization

Li Yang,Qinshu Liu,Jie Xu,Jing Hu,Tiecheng Song

DOI: https://doi.org/10.1109/VTCFall.2018.8690994

2018-01-01

Abstract:Radio frequency identification (RFID) is a kind of automatic identification technology which can be used for indoor positioning system. In this paper, an indoor RFID location algorithm is proposed based on nonlinear support vector regression (SVR) and particle swarm optimization (PSO). The algorithm uses SVR to construct the nonlinear mapping relation between received signal ength indication (RSSI) and distance between the tags and the readers. Based on nonlinear mapping relation, the nonlinear equation set is constructed. In addition, through applying PSO to optimize the objective function converted by nonlinear equation set, the coordinate position of target tag can be estimated. Finally, the simulation of this algorithm and some other algorithms is conducted. The experimental results show that SVR-PS algorithm is more efficient than the previous algorithms in terms of positioning accuracy and positioning stability.