SAR PRF-ambiguity resolving by range-Doppler domain normalized variance maximization

By exploiting the statistical property of synthetic aperture radar (SAR) range migration (RM) in the range compressed range-Doppler (RD) domain, we discovered that the 'averaged range image' of the range compressed RD domain, which is obtained by averaging the signal intensity over the RM curve, has the highest normalized variance (NV). Based on this fact, a novel method for resolving SAR pulse repetition frequency (PRF)-ambiguity in the RD domain is proposed in this paper. With high computational efficiency and easiness of being incorporated into existing SAR imaging algorithms, the proposed method is robust to scene contrast, and has little dependence on the ambiguous Doppler centroid estimation accuracy for large ambiguity number case. Experimental results with real spaceborne SAR data demonstrate its effectiveness.