A WiFi-aided reduced inertial sensors-based navigation system with fast embedded implementation of particle filtering

Abstract

Global Positioning System (GPS) accuracy deteriorates significantly in dense urban areas and it is almost unavailable inside buildings. Thus, an alternative accurate navigation system for such GPS-denied systems is of great importance. In this paper, the popular IEEE 802.11 WLAN (WiFi) is utilized along with a MEMS-based reduced inertial sensors system (RISS) to provide an accurate and smooth positioning system for wheeled vehicles inside buildings based on WiFi received signal strength (RSS). The WiFi/RISS integration is performed based on a fast version of Mixture Particle Filter (PF) which is a nonlinear non-Gaussian filtering algorithm that handles well the complex MEMS inertial sensors and WiFi stochastic nature. The proposed system was physically implemented on an embedded system on an OMAP 600 MHz processor board and tested on a mobile robot. Results show that drifts of RISS are greatly removed and the scattered noisy WiFi positioning is significantly smoothed. Experiments show that the integrated system can provide smooth indoor positioning of 2m accuracy 60% of time.