A method of space-based interferometric synthetic aperture radar measurements (InSAR and DInSAR techniques) is designed for obtaining of different types of altitude data for the Earth surface, i.e. terrain heights and surface displacements. Phase noise is an intrinsic random uncorrelated additive process, which complicates data processing in such systems. The experimental study of the linear non-adaptive phase noise filters are studied using the previously developed technique for experimental evaluation of the InSAR processing algorithms effectiveness. The accuracy characteristics for the filters applied to ALOS PALSAR data for areas with hilly and low-mountainous terrain are obtained, and it is shown, that one of the filters gives a slight better results (with less error value) under the equal computational complexity.