Passive Human Tracking based on IMM Kalman Filter using WiFi Signal

Recently, device-free indoor tracking system using WiFi signal has made significant progress. However, many studies have focused on how to high-accuracy parameter estimation and matching, ignoring the influence of localization and tracking algorithms on the results. In this paper, focusing on the irregular motion of the target, we design a passive localization and tracking system based on Interactive Multiple Model (IMM) Kalman filter. First, we use the traditional sage algorithm to extract the angle estimates about target motion and then realize target localization. Due to the irregularity of the target motion, we construct the several different motion states of the moving target and realizing continuous tracking using an IMM Kalman filter. The experimental results show that our proposed system can improve localization accuracy and achieve a median tracking error of 0.63 m.