As an emerging orbital system with flexibility and brand application prospects, the highly elliptical orbit synthetic aperture radar (HEO SAR) can achieve both a low orbit detailed survey and continuous earth surface observation in high orbit, which could be applied to marine reconnaissance and surveillance. However, due to its large eccentricity, two challenges have been faced in the signal processing of HEO SAR at present. The first challenge is that the traditional equivalent squint range model (ESRM) fails to accurately describe the entire range for the whole orbit period including the perigee, the apogee, and the squint subduction section. The second one is to exploit an efficient HEO SAR imaging algorithm in the squinted case which solves the problem that traditional imaging algorithm fails to achieve the focused imaging processing of HEO SAR during the entire orbit period. In this paper, a novel imaging algorithm for HEO SAR is presented. Firstly, the signal model based on the geometric configuration of the large elliptical orbit is established and the Doppler parameter characteristics of SAR are analyzed. Secondly, due to the particularity of Doppler parameters variation in the whole period of HEO, the equivalent velocity and equivalent squint angle used in MESRM can no longer be applied, a refined fourth-order equivalent squint range model(R4-ESRM) that is suitable for HEO SAR is developed by introducing fourth-order Doppler parameter into Modified ESRM (MESRM), which accurately reconstructs the range history of HEO SAR. Finally, a novel imaging algorithm combining azimuth resampling and time-frequency domain hybrid correlation based on R4-ESRM is derived. Simulation is performed to demonstrate the feasibility and validity of the presented algorithm and range model, showing that it achieves the precise phase compensation and well focusing.

Synthetic aperture radars (SARs) require a larger pulse repetition interval to achieve a greater range swath width, which is often addressed by designing non-uniform pulse transmission patterns. However, decoupling the pulse transmission pattern and SAR imaging algorithms in exiting works hinders the full utilization of the anti-aliasing capability of non-uniform transmission during the imaging procedure. To address this issue, this article proposes a joint optimization method of learnable pulse transmission pattern and SAR imaging algorithm, namely, JointSAR. Specifically, we optimize the pulse transmission intervals in a random parameter freezing way and develop a deep unfolding network based on approximate observation model. Simulation and measured data-based experiments verify that the proposed method achieves better SAR imaging quality and maintains wide-swath imaging capability in comparison with existing state-of-the-art methods.

Since multichannel sliding spotlight synthetic aperture radar (SAR) can simultaneously achieve high-resolution and wide-swath (HRWS) remote sensing observations, the large volume of data collected from multiple channels increases the burden on the data link, while the estimation and compensation of channel errors further complicate the signal processing. To reduce hardware costs and algorithm complexity while ensuring HRWS capabilities, this letter proposes a novel single-channel ultrahigh-resolution and wide-swath sliding spotlight SAR imaging approach based on a Poisson disk sampling strategy. First, the proposed method adopts a Poisson disk sampling strategy to collect nonuniform azimuth sampling echo data, which has well anti-aliasing effects for the sparsity-driven reconstruction. Second, a modified Chirp Scaling algorithm, which integrates a fourth-order range model (FORM4) to precisely account for the spaceborne SAR curved orbit, is presented to achieve full-aperture focusing of ultrahigh resolution sliding spotlight SAR data across the entire scene. Then, the complete imaging reconstruction process is given, which mainly involves the operation of Deramp preprocessing based on nonuniform sampling sliding spotlight SAR data, as well as threshold iterative processing. Simulation results powerfully demonstrate the capability of the proposed MCSA to achieve a 0.3m resolution over a 4km×4km swath in non-uniformly sampled sliding spotlight SAR, with a 20% reduction in data acquisitiion compared to traditional approaches.