A Kalman filter with speed constraints for WiFi-based indoor localization

This paper presents a constrained Kalman filter for WiFi-based indoor localization. The contribution of this work is to introduce constraints on the object speed and to provide a robust test determining when to apply these constraints. The proposed approach is appropriate when the indoor environment is split into several areas, each area having its own variable organization. It is also applicable to objects that could be span around, as for example barcode readers in a hand. The proposed technique is experimented with a robot and three devices, on five different journeys, in a 6000m
²
warehouse equipped with six Wi-Fi access points. The results highlights that the proposed approach provides a 17% improvement in the localization accuracy.