Sparse aperture inverse synthetic aperture radar imaging of manoeuvring targets with compensation of migration through range cells

Abstract

In inverse synthetic aperture radar (ISAR) imaging of a target with significant manoeuvres, severe migration through range cells (MTRCs) and time-varying Doppler usually involve in the echoed signal. Both of them may challenge the conventional motion compensation and imaging methods, which are usually based on the assumption of small rotational angle with short coherent time duration. Moreover, for a multi-functional ISAR, full aperture (FA) data collection might be unachievable because of the conflict with other radar activities, resulting in sparse aperture (SA) data. In this study, an ISAR imaging method of manoeuvring targets with SA data is proposed, where MTRC compensation and time-varying Doppler are well accounted. By using a sparse chirp-Fourier basis with first-order range-azimuth coupling term, an improved orthogonal match pursuit algorithm is implemented to solve the sparsity-driven optimisation for FA data reconstruction from SA-ISAR measurements. As a primary item, MTRC compensation is seamlessly embedded into the FA signal reconstruction for ISAR imaging of manoeuvring targets. Both simulated and real data are utilised to confirm the effectiveness of the proposed method.

Sparse aperture ISAR imaging algorithm based on adaptive filtering framework

“…The modern spectrum estimation method [3] used spectral estimation theory to approximate the complete target echo, such as using the all‐pole model to realise the matching process and interpolation process, so as to improve the azimuth resolution. The sparse signal reconstruction method [4–7] transformed the sparse aperture imaging problem into the sparse signal reconstruction problem by the sparsity of the target, and then the sparse optimisation algorithms were conducted to solve the problem. Compressive sensing (CS) theory points out that [4, 7] when the signal is sparse, the original signal can be accurately obtained by the reconstruction algorithm with a small number of observations.…”

ISAR imaging and cross‐range scaling of high‐speed manoeuvring target with complex motion via compressive sensing

“…To overcome this problem, we adopt sparse recovery algorithms (SRAs) that are based on compressive sensing (CS) theory [14, 15] for the reconstruction of ISAR images from the SA dataset. A current trend is the application of CS theory to various radar signal‐processing areas [16–20]. According to the CS theory, it is well known that the exact recovery of an unknown sparse signal can be achieved through limited measurements by solving a sparsity‐constrained optimisation problem.…”

“…The main contribution of this paper is to present a new framework for ISAR imaging and scaling by simultaneously implementing TMC, RMC, and CRS using CS theory. To reconstruct the pixel value of each pixel in the column vector of ISAR image, the conventional CS‐based methods mentioned above utilise direct sparse reconstruction by directly solving the corresponding optimisation problems during the process of sparse recovery as follows [14–18]: where is the norm of a vector, i.e. the number of non‐zero elements in a vector, is the measurement vector, and denotes the sensing‐dictionary matrix.…”

“…In [17], a sparsity‐based method has been reported, where a sparse metric is defined to iteratively estimate the sparse scatterer coefficients and phase error. In [18], Huang et al adopted CS to achieve a high‐resolution and well‐focused ISAR image using SA data . However, these methods require an estimation of all pixel values of associated with the size of ISAR image, which is computationally expensive and inefficient.…”

Polarimetric object‐level SAR imaging method with canonical scattering characterisation by exploiting joint sparsity

“…Sparse representation (SR) method has been applied successfully in solving inverse problems such as magnetoencephalography [8], direction‐of‐arrival estimation [9, 10], SAR imaging and feature extraction [11, 12]. In essence, the radar backscattering containing the geometries information of the target can be sparsely represented through the atoms of the redundant dictionary which incorporating prior information about expected scattering behaviour.…”