A sparse particle filter for indoor localization using mobile phones

In this paper, we propose a novel sparse particle filter applied for indoor pedestrian localization. Unlike the traditional particle filters that usually contain thousands of particles, this algorithm uses one or several particle sequences in location tracking. The particle is split to cover all the possible tracks when matching ambiguity occurs. This algorithm can significantly reduce the computation burden as well as energy consumption. Field experiments are conducted using mobile phones and the results show that the proposed method can perform localization with much higher efficiency, high reliability, and merely slight loss of accuracy.