Research On High-resolution Wide-Swath Imaging Technologies Of Azimuth Multiple Phase Center SAR

Synthetic aperture radar is capable of acquiring high resolution image of the earth’s surface for all-weather and all-time in two dimensions, thus has been an indispensable surveillance tool for remote sensing of the earth. Whether in civilian applications such as disaster monitoring and marine surveying and mapping or in military applications such as wide area surveillance and ship detection, etc., SAR has drawm intensive attentions and has been widely used. Limited by the inherent system constraint, spaceborne SAR can not meet the requirements of continuous high-resolution wide-swath(HRWS) remote sensing posed by modern space-to-ground missions. Adopting multiple receiving channels displaced in along track, the azimuth multi-phase-center(AMPC) SAR system can operates with a pulse repetition frequency(PRF) dramatically lower than the Doppler bandwidth, thus is capable of achieving continuous HRWS observing. With an eye to the future HRWS remote sensing, we focus on data processing and system performance analysis of AMPC SAR in the paper, where the signal modeling and reconstruction, system performance analysis, array error modeling and calibration, and the new operation mode exploiting problems are investigated. Major contributions and innovations in the paper are summarized as follows:Inherent principle and signal model of AMPC SAR are interpreted in Chapter 2. Basing on the analysis of the system merit factor and the minimum antenna area, we reveal that, AMPC SAR does not overcome the minimum antenna area constraint either, At the cost of system sensitivity, however, AMPC SAR realizes HRWS imaging by means of reallocating the antenna array. Basing on the analysis of spatial sampling characteristics, the PRF values leading to uniform spatial sampling are summarized, then, corresponding equivalent PRF resulting from displaced phase center antenna(DPCA) processing is derived.Signal reconstruction and performance analysis of AMPC SAR are studied in Chapter 3. Basing on the analysis of the distribution characteristics of clutter spectrum, modified signal model is established, where an arbitrary reconstruction Doppler bandwidth, rather than a value limited to integral multiples of the PRF as indicated in existed literatures, is considered. Benefiting from the modified signal model, conventional digital beamforming(DBF) methods provide stable reconstruction performance and work well for a wide range of PRF. Unified DBF processing frame is discussed. Then, reconstruction performance of the a series of processing methods are analyzed in the unified DBF processing frame, such as signal to noise ratio(SNR) and azimuth ambiguity to signal ratio(AASR). Theoretically, it is proved that the SNR performance resulting from LS method improves with the decreasing of the reconstructed Doppler bandwidth as well as with the increase of the receiving channel number.The issues on array error modeling and array error calibration are investigated in Chapter 4. Channel errors and position errors of the phase centers are modeled as array errors, including position errors in the along track, amplitude imbalances, and phase errors. it is revealed that the phase errors introduced by phase center errors are closely related to the incident angle, thus are range-variant. On ground of the attenuation characteristics of the clutter power spectrum, we propose a novel array error estimation method. The new method provides robust error estimation performance by minimizing the ratio between the integrated clutter power in the side Doppler zone and the integrated clutter power in the center Doppler zone. Orthogonal subspace(OS) method is modified to achieve robust error estimation performance in the case that only less receiving channels are available. Basing on the established array error model, we propose a model based phase fitting method. Considering the range-variant phase error introduced by phase center position error, we should estimating the phase errors for each range gate by adopting the conventional array error estimation algorithms, thus the processing can be computationally inefficient. The phase fitting method can acquire phase errors for every range gate basing on the phase errors obtained for only a few range gates. Thus, calibration of the range-variant phase errors can be more computationally inefficient.AMPC system based bi-directional SAR(Bi Di-SAR) imaging technologies are studied in Chapter 5. AMPC Bi Di-SAR observation mode is proposed to overcome the single-channel Bi Di-SAR deficiencies in system performance and design. Basing on in-depth spectral analysis, we propose to use LS based method to realize spectral separation, thus extends observing capability and design freedom of Bi Di-SAR dramtically. Analytical expressions of SNR and AASR of the after and fore images are derived.Correctness of the theoretical results and validity of the proposed data processing strategies are demonstrated by simulation results and the results of real measured data processing.