NEW TRENDS IN SAR TOMOGRAPHY F. Baselice, A. Budillon, G. Ferraioli, V. Pascazio, G. Schirinzi Università di Napoli Parthenope Dipartimento per le Tecnologie Centro Direzionale, Is. C4, 80143 Napoli, Italy A. Evangelista Università di Cassino DAEIMI Via Di Biasio, 43, 03043 Cassino, Italy 1. INTRODUCTION Three dimensional Synthetic Aperture Radar (SAR) imaging of earth surface has received a growing interest in recent years, thanks to the launch of new high-resolution radar sensors (TerraSAR-X and COSMO-SkyMed) which provide a large amount of images acquired with advanced SAR sensors. 3-D SAR image formation provides the scattering scene estimation along azimuth, range and elevation co-ordinates. It is based on multi-pass SAR data obtained, usually, by non-uniformly spaced acquisition orbits (see Fig 1). One of the main problems that have to be taken into account for the 3D SAR reconstruction concerns geometrical distortion. If we consider a ground height profile with three point scatterers (A, B and C) lying in the same range-azimuth resolution cell (see Fig 2), the acquired complex SAR signals related to the three scatterers collapse in the same resolution cell, producing the layover phoenomenon. The signal corresponding to the layover region depends both on amplitude (related to the material, roughness, viewing angle) and phase (related to the distance between sensor and object and speckle effects) of each single contribution involved. In this paper we analyze two different techniques to retrieve the height of the different contributions that collapse in a layover cell in order to achieve 3-D SAR imaging: Compressive Sensing and SAR Statistical Tomography. We compare the obtainable results in terms of multiple scatterers resolutions capabilities, number of baselines and signal to noise power ratio on simulated data in order to provide a set of instruments for the 3D SAR imaging able to tackle different scattering mechanisms in layover areas. 2. SAR STATISTICAL TOMOGRAPHY The first approach is based on SAR Statistical Tomography [1], which consists of trying to separate and distinctly estimate each complex contribute which collapse in layover pixels. In particular, it allows the joint estimation of both height and reflectivity of scatterers, providing the reconstruction of the height profile and of the scene reflectivity map. SAR Statistical Tomography is based on the assumption of a Gaussian distribution for the