Simultaneous ranging and self-positioning in unsynchronized wireless acoustic sensor networks

Abstract

Automatic ranging and self-positioning is a very desirable property in wireless acoustic sensor networks (WASNs) where nodes have at least one microphone and one loudspeaker. However, due to environmental noise, interference and multipath effects, audio-based ranging is a challenging task. This paper presents a fast ranging and positioning strategy that makes use of the correlation properties of pseudo-noise (PN) sequences for estimating simultaneously relative time-of-arrivals (TOAs) from multiple acoustic nodes. To this end, a proper test signal design adapted to the acoustic node transducers is proposed. In addition, a novel self-interference reduction method and a peak matching algorithm are introduced, allowing for increased accuracy in indoor environments. Synchronization issues are removed by following a BeepBeep strategy, providing range estimates that are converted to absolute node positions by means of multidimensional scaling (MDS). The proposed approach is evaluated both with simulated and real experiments under different acoustical conditions. The results using a real network of smartphones and laptops confirm the validity of the proposed approach, reaching an average ranging accuracy below 1 centimeter.