Personal Area Networks (PAN) are key topologies in pervasive Internet of Things (IoT) localization applications. In the numerous object localization techniques, centralization and synchronization between the elements are assumed. In this paper, we leverage crowdsourcing from multiple fixed and mobile elements to enhance object localization. A cooperative crowdsourcing scheme is proposed to localize mobile low power tags using distributed and mobile/fixed readers for GPS assisted environments (i.e., outdoor) and fixed readers for indoors. We propose Inertial-Based Shifting and Trilateration (IBST) technique to provide an accurate reckoning of the absolute location of mobile tags. The novelty in our technique is its capability to estimate tag locations even when the tag is not covered by three readers to perform trilateration. In addition, IBST provides scalability since no processing is required by the low power tags. IBST technique is validated through extensive simulations using MATLAB. Simulation results show that IBST consistently estimates location, while other indoor localization solutions fail to provide such estimates as the state-of-the-art techniques require localization data to be available simultaneously to provide location estimation.
Keywords: RSSI; crowdsourcing; inertial sensor; localization.