On extended source localization in multibaseline and multifrequency SAR interferometry

Baseline or frequency diversity have been recently proposed to reduce problems of interferometric phase ambiguity and data noise in synthetic aperture radar interferometry (InSAR) for topographic mapping. This paper presents a general model-based multibaseline multifrequency (MBMF) InSAR framework. A statistical model for the MBMF SAR-processed echoes from extended natural targets is presented, then the Cramèr-Rao lower bound on the MBMF-InSAR phase accuracy is derived. The maximum likelihood estimator for MBMF-Insar is introduced, and its simulated performance is compared to the bound for the common cases of dual baseline InSAR, dual frequency InSAR, and for a novel joint dual baseline dual frequency InSAR configuration. Performance of these configurations and improvement over conventional interferometry is analyzed.