The Research Of Resolution Enhance For Synthetic Aperture Radar

Synthetic aperture radar(SAR) has the characteristics of all-day, all-weather and high resolution imaging. It is an important goal to improve resolution of SAR imaging because the conventional SAR imaging algorithm is limited in resolution. This paper researches the algorithm of high-resolution SAR imaging, and the effectiveness of the algorithm are verified through the simulation experiments. The main work includes:1. Range cell migration is introduced, based on which original range Doppler algorithm(RD) method and the linear range migration correction RD method are given in this paper. Hence, their advantages and limitations are discussed. Moreover, the feasibility of the algorithm is verified by means of point-scatter target. Finally, we introduce several kinds of frequently-used imaging algorithm, such as CS, NCS, etc. And the process of the algorithm is given in detail.2. Put forward a method to enhance resolution by extrapolating the bandwidth of the spatial frequency spectrum. We combine the target polarization information with the spectral extrapolation technology: firstly, full-polarization data is decomposed according to Pauli’s basis; Then deduces the spectrum extrapolation method in the mode of full-polarization using the redundancy and complementarity between the linear prediction model and the target of different polarization channels on different data. Finally, under this decomposition, we develop Pol SAR bandwidth extrapolation algorithm, which is jointly extrapolating the frequency of all polarization-channel data. Compared with the traditional spectrum extrapolation method, due to adopt the joint extrapolation method, we can obtain higher resolution.3.The resolution improves with the increasing bandwidth, however the ultra-wideband signal is difficult to achieve due to its high requirement in hardware. Thus, this paper proposes an ultra-wideband signal method by coherently combining multi-bands measurement. First, we use singular value decomposition model to estimate the subband order number, based on which the signal can be decomposed into signal subspace and noise subspace. Furthermore, we estimate the subband signals by using the root-MUSIC algorithm. Hence, we estimate the all-pole model coefficient of the subsidiary zone using linear least-square iterative, and adjust the subsidiary zone. Non-linear least-square iterative is employed to obtain global all-pole signal model. Finally, the effectiveness of the proposed method is applied into SAR imaging to verify its resolution improvement by means of simulated targets.