This paper presents a consistent adaptive framework for high resolution indoor positioning system based on signal strength of the popular IEEE 802.11 WLANs (known as WiFi). The proposed system is designed based on service-oriented-architecture where each module is encapsulated into a service and services can communicate between each other using any communication protocol such as TCP IP or HTTP. This design gives the system the flexibility to be implemented on a variety of configurations to serve wide spectrum of WLAN-enabled devices especially smart-phones. In addition, the system constitutes an adaptive framework built on the concept of mutual signal strength observations between access points. The mutual signal strength observations between access points are broadcasted and monitored by different services in the system. Having the access points locations known by another service, the monitored signal strength information is used to periodically build up-to-date models that can predict signal strength anywhere anytime in the environment using a combination of path loss formulas and Gaussian Process Regression. Thus, without any human interference or offline data collection, the system is continuously aware about the signal strength distribution in the environment. This has the advantage of autonomously adapting to long-term environment changes such as restructuring the floors or changing walls surfaces. For the short-term changes in signal strength, a fast clustering-based anomaly rejection algorithm is developed to reduce this noise effects. The system was setup and tested in Trusted Positioning office inside the Calgary Technology Centre in Calgary, Alberta, Canada. Android smartphones with floor map of the area was used as positioning devices. Results showed that the positioning accuracy ranges from 1 to 2 meters most of the time with some few outliers which nominate the system to be a promising alternate positioning system for enterprises and other indoor environments that doesn't have access to Global Navigation Satellite Systems.

Additional Metadata
Conference Institute of Navigation International Technical Meeting 2013, ITM 2013
Citation
Atia, M, Georgy, J. (Jacques), & Noureldin, A. (Aboelmagd). (2013). An enterprise service oriented architecture-based high resolution WiFi indoor positioning system. In Institute of Navigation International Technical Meeting 2013, ITM 2013 (pp. 752–757).