PRF-ambiguity resolving by wavelength diversity

For high-precision synthetic aperture radar (SAR) processing, the determination of the Doppler centroid is indispensable. The Doppler frequency estimated from azimuth spectra, however, suffers from the fact that the data are sampled with the pulse repetition frequency (PRF) and an ambiguity about the correct PRF band remains. A novel algorithm to resolve this ambiguity is proposed. It uses the fact that the Doppler centroid depends linearly on the transmitted radar frequency for a given antenna squint angle. This dependence is not subject to PRF ambiguities. It can be measured by Fourier-transforming the SAR data in the range direction and estimating the Doppler centroid at each range frequency. The achievable accuracy is derived theoretically and verified with Seasat data of different scene content. The algorithm works best with low contrast scenes, where the conventional look correlation technique fails. It needs no iterative processing of the SAR data and causes only low computational load