Atmospheric compensation of acoustic signatures of aircraft

Abstract

The U.S. Army Research Laboratory is investigating using low size, weight, power, and cost (SWaP-C) acoustic sensors to classify unmanned aircraft systems (UAS). One issue complicating UAS classification using acoustics is that the atmosphere attenuates the signature of the targets in a complex manner. To address this issue, we developed and tested a technique to mitigate the effect of atmospheric attenuation. We used Bass's model with inputs from temperature, relative humidity and range to estimate atmospheric attenuation, then we used techniques based upon a Wiener filter implemented in the frequency domain to adjust the amplitude of the measured signal. The Wiener filter reduced the amplification of the noise while improving the reproducibility of the signature at different ranges, particularly at higher frequencies. The results suggest that preprocessing the data using this technique should improve the performance of acoustic classifier algorithms.