A Ubiquitous Indoor–Outdoor Detection and Localization Framework for Smartphone Users

There are significant research efforts on indoor positioning technologies as well as GPS-based outdoor localization and navigation. So, there is a requirement for effectively integrating the indoor and outdoor localization services and load the appropriate system depending on the context. Indoor/outdoor context detection can be done ubiquitously through exploiting the smartphone sensors that play the role of interconnecting the two services together. Thus, our contribution in this paper is to propose a ubiquitous indoor/outdoor localization framework that can not only detect the context but can also localize at a finer granularity around such transitional areas, such as the building gates. A combination of sensors is needed for precise positioning around that area. To implement the framework, a data collection application is built for smartphones. Experiments are also conducted based on data collected for several use cases in the university campus. The results indicate that a classification accuracy of 88% could be achieved for indoor–outdoor detection while the average localization error is around 1 m.