Dual-baseline SAR interferometry from correlated phase signals

In this paper we consider the problem of the estimation of the digital elevation model (DEM) of a ground scene starting from dual baseline interferometric synthetic aperture radar (SAR) signals. classical maximum likelihood (ML) estimation techniques require the knowledge of the joint probability density function (PDF) of the two measured interferometric phases. Usually the joint PDF expression is derived assuming that the two interferograms are statistically independent. This assumption is not verified for dual baseline configurations based on three image acquisitions. In this paper we compute the dual baseline joint pdf of the interferometric phases in a closed form, taking into account the mutual correlation of all the involved SAR images and without using any approximation. This expression can be exploited for the computation of the maximum likelihood function and for the method accuracy assessment by means of the Cramer Rao lower bound (CRLB) evaluation.