A Modified Hybrid PDA/FIR Localization Algorithm for Wireless Sensor Network

In indoor localization based on wireless sensor networks, due to the complex and changeable indoor environment, the propagation of wireless signals will be interfered by obstacles and walls to produce errors, which is called non-line-of-sight (NLOS) error. Although the probability data association (PDA) filtering algorithm is one of the most successful positioning algorithms in line-of-sight, the localization accuracy will be reduced and may lead to localization failure in the case of NLOS. To overcome this question, a modified hybrid filtering algorithm combining finite impulse response (FIR) filter and PDA filter is proposed. The FIR filter is used to recover the failed PDA filter, so as to alleviate the influence of NLOS on PDA filter location algorithm. The proposed FIR/PDA (FPDA) hybrid filter uses PDA filter and FIR filter as main filter and auxiliary filter respectively. Only when the PDA filter diverges, the auxiliary filter is enabled to generate state estimation and estimation error covariance and reset the main filter. This algorithm combines PDA filter with FIR filter to improve algorithm reliability and stability under NLOS conditions. The simulation and experiment results show that FPDA is superior to MPDA, MJPDA, EKF and other algorithms in robustness and localization accuracy.