Context Determination for Adaptive Navigation using Multiple Sensors on a Smartphone

Han Gao, Paul D. Groves

Abstract: Navigation and positioning is inherently dependent on the context, which comprises both the operating environment and the behaviour of the host vehicle or user. No single technique is capable of providing reliable and accurate positioning in all contexts. In order to operate reliably across different contexts, a multi-sensor navigation system is required to detect its operating context and reconfigure the techniques accordingly. This paper aims to determine the behavioural and environmental contexts together, building the foundation of a context-adaptive navigation system. Both behavioural and environmental context detection results are presented. A hierarchical behavioural recognition scheme is proposed, within which the broad classes of human activities and vehicle motions are detected using measurements from accelerometers, gyroscopes, magnetometers and the barometer on a smartphone by decision trees (DT) and Relevance Vector Machines (RVM). The detection results are further improved by behavioural connectivity. Environmental contexts (e.g., indoor and outdoor) are detected from GNSS measurements using a hidden Markov model. The paper also investigates context association in order to further improve the reliability of context determination. Practical test results demonstrate improvements of environment detection in context determination.
Published in: Proceedings of the 29th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2016)
September 12 - 16, 2016
Oregon Convention Center
Portland, Oregon
Pages: 742 - 756
Cite this article: Gao, Han, Groves, Paul D., "Context Determination for Adaptive Navigation using Multiple Sensors on a Smartphone," Proceedings of the 29th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2016), Portland, Oregon, September 2016, pp. 742-756. https://doi.org/10.33012/2016.14573
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In