The thesis consists of two parts. The first part is devoted to the theory of dual-orthogonal polarimetric radar signals with continuous waveforms. The thesis presents a comparison of the signal compression techniques, namely correlation and de-ramping methods, for the dual-orthogonal sophisticated signals. The novel time-frequency representation of beat signals in frequency modulated continuous wave (FM-CW) polarimetric radar with simultaneous measurement of scattering matrix elements is shown. Since the sounding sophisticated signals usually have large time-bandwidth product, the observed object motion can result in specific bandwidth effects in the scattered signals. A novel wide-band correlation processing for overcoming the limitations due to these bandwidth effects is proposed for polarimetric radar signals. Also a novel wideband model of the de-ramped signals is described. The second part of the thesis is devoted to advanced processing in polarimetric radar with continuous waveforms (and focus on polarimetric FM-CW radar). A novel technique for continuous “quasi-simultaneous” measurement of the elements of the scattering matrix, which can provide high isolation level between the branches in the radar receiver channels, has been proposed in this PhD thesis. The ambition to increase the radar performance, namely to improve the radar range resolution, has led to the development of a novel flexible de-ramping processing applicable in single-channel and in polarimetric FM-CW radar. The problem of isolation in the polarimetric FM-CW radar receiver is especially acute. A novel method allowing to mitigate the problem is developed. The modeling results of the proposed method show its high efficiency.