Abstract : Airborne broadband ultrasound (ABU) covers a wide stretch of inaudible acoustic frequencies which have been used to provide scalable (i.e. multiuser) and highly robust ranging and localization of static tags in indoor environments. This is useful for mobile computing applications and robotics. However, because ABU signals are wideband and propagate relatively slowly, they are prone to aggressive Doppler distortion as a result of human-scale movements indoors. Additionally, the ABU channel is characterized by pronounced distance-dependent variations. This paper gives a treatment of the ABU channel both theoretically and experimentally, in order to inform the design of ABU tracking systems. Then, Doppler-tolerant ABU tracking algorithms are developed. Their performance is reported, using a direct-sequence (DS) code-division multiple access (CDMA) signalling approach. Successful motion inference is demonstrated in the ABU band.
