INS/Wi-Fi Based Indoor Navigation Using Adaptive Kalman Filtering and Vehicle Constraints Wennan Chai * , Cheng Chen ** , Ezzaldeen Edwan * , Jieying Zhang * and Otmar Loffeld * * Center for Sensorsystems (ZESS), University of Siegen, Siegen, Germany, Email: {chai, edwan, czhang, loffeld}@zess.uni-siegen.de ** Institute of automatic control engineering (RST), University of Siegen, Siegen, Germany, Email: c.chen@ zess.uni- siegen.de Abstract—Due to the complementary nature of inertial navigation system (INS) and Wi-Fi positioning principles, an INS/Wi-Fi integrated system is expected to form a low- cost and continuous indoor navigation solution with better performance than using the standalone systems. In this paper, we explore the integration of Wi-Fi measurements with data from microelectromechanical systems (MEMS) based inertial measurement unit (IMU) for indoor vehicle navigation. Two enhancements, which employ adaptive Kalman filtering (AKF) and vehicle constraints, for supporting the integrated system are presented. One field experiment has been conducted for estimating the trajectory of a mobile robot vehicle. The numerical results show that the enhanced integrated system provides higher navigation accuracy, compared to using standalone Wi-Fi positioning and conventional INS/Wi-Fi integration. Index Terms Adaptive Kalman Filter; Fingerprinting; Inertial Navigation System; Vehicle Constraints Wireless Local Area Networks I. INTRODUCTION For outdoor positioning and navigation, solutions based on Global Navigation Satellite System (GNSS) are satisfactory in most applications. But such technology is not utilizable for most indoor applications. Therefore, other positioning techniques have been developed for indoor environments lately, e.g., the methods based on Wireless Local Area Networks (WLAN), Bluetooth, Radio Frequency Identification (RFID), Ultra Wideband (UWB), infrared and ultrasound, etc. Among these techniques, the approach on the basis of exploiting 802.11 WLAN (Wi-Fi) is attractive and is expected to yield a cost-effective and easy-accessible solution. The localization methodologies based on Wi-Fi rely on the signal to noise ratio (SNR) or the received signal strength (RSS) [19]. Most of them, comprising the widely referred RADAR method, employ fingerprinting methods [4]. The localization accuracy using fingerprinting depends on the database density. Besides, errors and discontinuities in the trajectory of the object are introduced by the Wi-Fi signal fluctuations. The MEMS based IMU is widely used in navigation applications. The IMU provides the motion information of the object with a high update rate and it can achieve high precision in short time duration. However, without external aiding, the system suffers from local anomalies and the error drifts quickly with time [4]. Therefore, in outdoor environments, the IMU is often integrated with Global Positioning System (GPS) receiver. The integration of INS and GPS has been proven to be a reliable solution for continuous outdoor navigation [6][10]. The combination of INS and Wi-Fi positioning can also yield a synergetic effect resulting in higher navigation performance [12]. One widely employed algorithm for the fusion of the both systems is Kalman filtering (KF) [6]. In [14], a Wi-Fi assisted dead-reckoning navigation system using extended Kalman filtering (EKF) is presented and the experimental results are encouraging. The performance of the INS/Wi-Fi integrated system can be further improved without hardware changes. In this paper, one enhanced system is presented, which makes use of AKF and vehicle constraints. In order to obtain the highly accurate estimates from the KF algorithm, the prior statistical information about the system process and measurement noises should be estimated. AKF is proposed for adapting the measurement and process noise matrices in a Kalman filter. In [3] and [9], the AKF was used for integrating navigation grade INS measurements with GPS and the results showed the improvements in navigation performance. Incorporation of vehicle constraints into the estimation process has been recently proposed as a means to avoid the use of external sensors. In [13] and [22], vehicle constraints has been used for supporting INS navigation in urban environments and the outperformance of the constraint aiding system has been proven with experiment results. Based on above researches, the AKF and vehicle constraints are expected to improve the INS/Wi-Fi integration for indoor vehicle navigation. The rest of this paper is organized as follows. The second section describes the INS process model, the Wi-Fi fingerprinting approach and the unscented Kalman filter (UKF) employed for the integrated system. The third section presents enhancements using AKF and vehicle constrains. In the fourth section, a field experiment is described and the numerical results are analyzed. Finally, conclusions of this research are provided. 978-1-4673-1439-8/12/$31.00 ©2012 IEEE 36