Wi-Fi Indoor Localisation Based on Collaborative Ranging Between Mobile Users

H Jing, J. Pinchin, C. Hill, T. Moore

Abstract:	Location-based applications have gained a growing popularity in everyday application for both military and civil usage over the last decade as Global Navigation Satellite System (GNSS) technology matures. However, as both users and application extend to a much broader scope than what GNSS was initially designed for, applications in GNSS-denied areas have drawn a great deal of attention from researchers. Wi-Fi fingerprinting has become a popular solution when faced with positioning and navigation problems in complex indoor environments. Yet the accuracy of fingerprinting is rather limited and the system is prone to destruction if the structure of the building or Wi-Fi access point network changes. However, if a number of mobile users share each of their signal and positioning information as well as ranging data between other users collaboratively to form a local ad-hoc network, the information could be used to correct any failure in the fingerprinting process and provide more signal information to give a robust positioning result. This paper implements collaborative fingerprint positioning using Particle Filters which integrates multi-positioning-sensor information from available users. The filter takes a coarse dead reckoning model as the prediction model and takes into account the uncertainty of fingerprint positioning results, then outputs a series of possible position solutions as a measurement input for updating particle probabilities. Ranging data and each user’s Wi-Fi Received Signal Strength (RSS) map is then shared and included in the filter to restrict particles from drifting in the wrong direction. The algorithm is tested with simulated data based on the performance of collected Wi-Fi and Ultra-wide Band (UWB) signals in indoor environments. After which, experiment is carried out using real data collected inside Nottingham Geospatial Building. A Wi-Fi database is setup by walking along the corridors with a MicroStrain inertial sensor attached onto the user’s foot to provide positioning information while the user carries a laptop to scan Wi-Fi RSS data inside the building. ZUPT correction corrects the collected data which is then processed through a particle filter to achieve the best optimised inertial measurement positions. During the positioning phase, coarse inertial measurements, Wi-Fi data and ranging data from UWB for two users are collected in simple geometric environments. Data are post-processed using the proposed algorithm to obtain positioning results. Simulation test results have shown improvement in positioning accuracy. By implementing collaborative ranging between users, the system becomes less reliant to Wi-Fi network. Real Wi-Fi signals contain much more noise but positioning result improved situations are been recognized and analysed to develop the filter into a more robust one. Current indoor positioning technologies mostly rely on pre-installed infrastructure, therefore cannot work in a new environment. By using this method, the user could potentially start off the positioning phase without Wi-Fi RSS database and only starts to scan and collect RSS as the user enters the area of interest. Therefore the training and positioning phase is carried out at the same time. Useful information could be added into the system and shared among other users during positioning. Future work aims to eliminate failure situations as much as possible by adding more users into the collaborative system. Further research on taking out infrastructure information is also looked into. The model would be placed in a more complicated environment which is close to real-life situations to analyse the robustness of the algorithm.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	1317 - 1324
Cite this article:	Jing, H, Pinchin, J., Hill, C., Moore, T., "Wi-Fi Indoor Localisation Based on Collaborative Ranging Between Mobile Users," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 1317-1324.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In