Radar signature extrapolation for FISC

We present a frequency extrapolation scheme to speed up the signature prediction procedure using FISC by avoiding exhaustive computations. We adopt a model-based approach to the frequency extrapolation problem. The key in the success of an extrapolation algorithm is a good model of the physical observable to be extrapolated. We propose a scheme that parameterizes the current on the target, which is available in the FISC solution, based on a multipath excitation model. We first run FISC for the scattering problem at several frequency points, then apply the ESPRIT superresolution algorithm directly to the induced current output at these frequency points to extract the time-of-arrival and amplitude parameters. Once these parameters are obtained, the frequency dependent model of the induced current on each facet of the target can be constructed. Thus the induced currents over the whole frequency band can be extrapolated and the multifrequency far field can be calculated. The range profiles of a low-Q missile model are calculated as an example to demonstrate the performance of the algorithm.