The vortex electromagnetic (EM) wave carrying orbital angular momentum (OAM) has shown significant potential in high-resolution radar imaging. However, the Bessel function modulation (BFM) effect caused by the vortex wave’s physical nature influences the imaging performance, especially under the limited OAM modes. To address this issue, we first establish a computationally efficient 2-D vortex radar imaging model and analyze the influence of the BFM effect on vortex radar imaging from the point spread function (PSF). Then, we propose a novel EM vortex radar imaging algorithm based on joint low-rank and sparse constrained representation. Specifically, the vortex radar echo data exhibits the inherent low-rank property, and hence a low-rank regularization term is introduced to remove the BFM effect. Moreover, a constraint corresponding to the image sparsity is adopted to eliminate the interference caused by the limited OAM modes. Hence, the high-resolution vortex radar imaging problem is modeled as a double-constrained underdetermined optimization problem. An iterative algorithm based on the alternating direction method of multipliers (ADMM) framework is further designed to solve the problem with high efficiency. Experimental results based on synthetic, EM, and measured data validate the effectiveness of the proposed algorithm.
31 December 2024 in IEEE Transactions on Aerospace and Electronic Systems
The vortex wave featuring specific wavefronts and orthogonal orbital angular momentum (OAM) modes has demonstrated broad prospects in radar forward-looking imaging applications. However, the Bessel function modulation effect arising from the physical property of vortex waves and the limited OAM modes affect the imaging performance. To solve this issue, we present an innovative vortex wavefront modulated radar (VWMR) forward-looking imaging method by leveraging the underlying continuity of the scatterers within the target scene to achieve improved VWMR imaging reconstructions through a hierarchical Bayesian framework. An extended two-dimensional structural spike-and-slab hierarchical Bayesian prior model is proposed to statistically promote spatial continuity of the target scene, where the support and the scattering coefficient values are both correlated with their neighbors. Specifically, the coefficient values exhibit correlation through a soft coupling mechanism that shares hyperparameters among neighboring coefficients. The support of the target area is enforced to be dependent on its immediate neighbors, further promoting the zero or nonzero scatterers to gather in a spatial location-correlation behavior. The variational Bayesian expectation-maximization method is exploited for the approximate posterior inference of the latent variables and the estimation of the model parameters. Comprehensive experimental results using synthetic, electromagnetic, and measured data validate that the proposed method offers superior reconstruction performance over other reported VWMR imaging algorithms.
The generalized sparse iterative covariance-based estimation ($q$-SPICE) algorithm has significant capability to enhance angular resolution in scanning radar applications. However, two drawbacks remain: one is that it focuses solely on the azimuth dimension and ignores the pitch dimension, and the other is that its computational complexity grows exponentially with increasing data size. To address these, a 2D weighted $q$-SPICE (W$q$-SPICE) is proposed in this letter, which enables rapid azimuth-pitch sparse reconstruction of the scanning radar. The proposed algorithm is an extension of $q$-SPICE within the framework of weighted SPICE (WSPICE), and it is also a generalized version of the recent 2D WSPICE algorithm. Moreover, a selection criterion for $q$ is provided. Experiment results based on simulated and real radar data illustrate the superiority of the proposed method over its counterparts in the task of 2D resolution enhancement for the scanning radar.