One of the most critical stages in height reconstruction from Synthetic Aperture Radar (SAR) interferometry is the two-dimensional phase unwrapping stage. The observables, i.e. the gathered complex radar data, are characterised by a measurable phase between 0 and 2, whereas the true phase of each sample, directly related to the propagation path of the electromagnetic wave and therefore to the sensor-to-target distance, is necessary for effective derivation of the altimetric information embedded in the illuminated scene. It is possible to derive absolute phase information by exploiting a priori phase information of some ground control point, whose position is completely known in the scene, even if most of the times there are phase inconsistencies, i.e. phase differences between adjacent samples greater than . This work presents a new strategy for effective phase unwrapping in the presence of inconsistencies, defining an automatic and effective algorithm for correct reconstruction of the absolute phase of the image, even when phase noise and discontinuities can impair other algorithms. Simulated interferometric SAR images will be used to validate our technique and to compare with other methodologies proposed by other authors. Experimental results on simulated images show the robustness of our methodology, even in the case of low-coherence images.