A signal processing model is presented for acoustic sensors on ground and unmanned aerial vehicles
(UAV). Such sensors normally experience more flow noise than stationary sensors, because moving
platforms must vary their velocity to accomplish their missions. In the case of the UAV, this includes
sufficient speed to remain airborne. Unfortunately, high airflow speeds over the sensor cause turbulence
noise that tends to confound the acoustic detection of signals from sources of interest on the ground. This
model transforms the fluctuations in the magnitudes and the phase angles of signals and turbulence noise.
The temporal coherences of the signals are improved to the point where detections can be made
unambiguously, and be based on temporal coherence rather than on the signal-to-noise ratio, which is the
customary way to detect signals. Additionally, because the flow noise is temporally incoherent, it is easily
discriminated against. The model transforms phase and amplitude fluctuations in a such a manner that the
temporal coherences of the signals are increased. This makes them more easily exploited to achieve signal
processing gains, such as increases in signal-to-noise ratio and automatic detection. The rationale for this
model is that both signal and noise posses magnitude, but only signals posses temporal coherence. Two
transformations are presented herein. One transforms the phase angles, and the other one transforms the
spectral amplitudes. The transformations give the amplitudes and phase angles similar exploitable
coherence characteristics, while the corresponding noise incoherence is easily attenuated.