Enhancing Localization Scalability and Accuracy via Opportunistic Sensing

Abstract

Using a mobile phone for fine-grained indoor localization remains an open problem. Low-complexity approaches without infrastructure have not achieved accurate and reliable results due to various restrictions. Existing accurate solutions rely on dense anchor nodes for infrastructure and are therefore inconvenient and cumbersome. The problem of beacon signal blockage further reduces the effective coverage. In this paper, we investigate the problems associated with improving localization scalability and accuracy of a mobile phone via opportunistic anchor sensing, a new sensing paradigm which leverages opportunistically connected anchors. One key motivation is that the scalability of the infrastructure-based localization system can be improved by lifting the minimum requirement for anchor numbers or constellations in trilateration. At the same time, location accuracy under insufficient anchor coverage will be improved by exploring the opportunity of diverse data types rather than deploying more anchor nodes. To enable this highly scalable and accurate design, we leverage low-coupling hybrid ranging using our low-cost anchor nodes with centimeter-level relative distance estimation. Activity patterns extracted in users’ smartphones are utilized for displacement compensation and direction estimation. The system also scales to finer location resolution when anchor access is improved. We introduce robust delay-constraint semidefinite programming in location estimation to realize optimized system scalability and resolution flexibility. We conduct extensive experiments in various scenarios. Compared with existing approaches, opportunistic sensing could improve the location accuracy and scalability, as well as robustness, under various anchor accessibilities.