Since UWB-SAR systems operate in a frequency band populated by TV broadcasting stations, Radio Frequency Interference (RFI) is a major problem. The development of successful RFI suppression techniques for UWB radars is crucial for their operation in many geographical areas. In this proposal we outline several advanced RFI suppression techniques for UWB-SAR, which will be developed and evaluated during the Phase I project. We consider three specific approaches. The first involves modification of the existing class of RFI suppression techniques which use a time-invariant transversal filter. We propose several ways to modify the computation of the filter coefficients, to provide better control of the filter response and minimize signal distortion. The second is based on the application of advanced time-frequency analysis to the collected SAR data prior to image formation, to identify/detect features of the received signal which are related to RFI, and then remove/suppress these features using a time-varying filter. The third is a relatively new approach which seems to be quite promising for processing non-stationary signals. This approach, which involves nonlinear processing is based on the recently developed idea of wavelet de-noising, which offers a new way for separating a signal into components which have different characteristics. Our objective is to develop RFI suppression algorithms based on each of these approaches, evaluate their performance, and assess their cost/benefit tradeoffs.
