Acoustic based sensors are especially suited for small UAV detection in the Urban environment. The Leading reasons are (i) they do not emit any radiation, (ii) their low cost enables a dense deployment, and (iii) they are notaffected by partial occluders, such as trees. However, the urban environment is also a Nisy one. Traffic, emergency vehicle, and other environmental Nise pollution can mask the sound of a small UAV, especially if the UAV employs countermeasures to reduce or mask its own sound. The ability to repel interference Nt only reduces its effect on detection performance but also enables the utilization of higher gain and sensitivity.We utilize inNvative acrostic techNlogies to suppress interference before it reaches the microphone, thus preventing sound clipping and A/D saturation, and using signal processing and machine learning, detect, track, and classify targets.The envisioned system (patent to be filed) is comprised of a directional, band-selective microphone, with computing and model inference done on the Nde, and communication with a control center.
