Modeling of FM broadcast signals for use in bistatic radar applications
The purpose of this project is threefold: To determine the optimum range-Doppler characteristics of an FM broadcast, to characterize a typical FM broadcast signal in terms of the properties of the ambiguity function, and to develop a model for the FM broadcast signal as a band-limited multi-tone FM stochastic process. The way to achieve these objectives is to analyze the response of a matched filter at the receiver by considering two modulating input signals. The inputs are a band-limited white Gaussian noise and a segment of a typical broadcast. The analysis of these signals is performed in the range-Doppler domain by observing the behavior of the ambiguity function. The results of the analysis show that the band-limited white Gaussian noise yields an optimum ambiguity function with a narrow mainlobe and low sidelobes evenly distributed over the range-Doppler plane. In contrast, the ambiguity function of a typical broadcast will exhibit a wider mainlobe and higher sidelobes. By adjusting parameters of the multi-tone FM stochastic process, a good match to the observed characteristics of an actual FM broadcast can be obtained. Thus, the signal model can be used to estimate the limitations in the detection of moving targets by means of a radar system that exploits FM signals of opportunity.
Sullivan, Rebecca, "Modeling of FM broadcast signals for use in bistatic radar applications" (2004). ETD Collection for University of Texas, El Paso. AAI1423722.