Anti-multipath Indoor Direction Finding Using Acoustic Signal Via Smartphones.

Xiaopu Wang,Yan Xiong,Wenchao Huang
DOI: https://doi.org/10.1007/978-3-319-22047-5_11
2015-01-01
Abstract:Direction Finding plays a significant role in indoor localization research. We introduce a novel scheme of anti-multipath indoor direction finding via smartphones. This scheme does not rely on any fingerprints or specialized devices. Users only need to move their smartphones over a short distance for finding the anchor points (ordinary speakers) we preset in the scene. Due to the intricate indoor environment, the multipath effect has pronounced negative influence to the direction finding accuracy. Our work is aimed at mitigating the adverse impact of multipath effect which is based on MIMO (multiple-input, multiple-output) technique. However, processing the signals received on an array of sensors is the pith of MIMO. As far as we know, the directional antenna is a specialized device which is expensive and bulky. Hence, we propose a method of sampling and processing high frequency acoustic signals to emulate a virtual antenna array. In this case, we leverage smoothed MUSIC (Multiple Signal Classification) algorithm to estimate the DOA (direction of arrival) of the target signals. We design and deploy our system under various circumstances in an underground parking lot. By getting rid of the multipath effect interference, the extensive experiments show that our system of the direction finding are precise and high-efficiency in daily use.
What problem does this paper attempt to address?