An Approximate Least-Squares Method for Fast Gridless Radar Imaging

This paper presents a fast method called approximate least-squares (ALS) algorithm to address the off-grid problem in tomographic synthetic aperture radar (TomoSAR) imaging. First, we give the model of TomoSAR imaging and formulate the off-grid optimization problem. Next, we propose ALS algorithm to solve the optimization problem. ALS uses the second-order approximation of Taylor expansion to represent the off-grid optimization problem, and gives approximate closed-form solution for fast processing. Finally, numerical simulations confirm that ALS achieves gridless TomoSAR imaging faster while ensuring performance.