Modified two-stage separated virtual steering vector-based algorithm for high resolution inverse synthetic aperture radar imaging

A computationally efficient propagator and two-stage separated virtual steering vector based algorithm (SVSV) is recently developed by D. Liu and J. Liang for two-dimensional (2D) frequency estimation. SVSV has introduced a novel electric angle z to jointly estimate the 2D components (frequencies or scatterers locations) and to avoid the incorrect mismatch. However, SVSV cannot resolve the multiple 2D components having equivalent value of z. This problem limits the use of SVSV in ISAR imaging application. Hence, in this paper, we have proposed a new modified SVSV method called MSVSV for 2D ISAR imaging. Main contributions of our work are: (a) use of the propagator based subspace methods (SVSV and MSVSV) for 2D high resolution ISAR imaging. (b) We have suggested a technique to estimate the number of unknown 2D components prior to calculating the propagator operator from the covariance matrix. (c) We have incorporated a technique, due to which MSVSV can successfully resolve the multiple scatterers sharing the same electric angles, where SVSV fails. Radar images are generated for simulated and experimental target data (collected at AERO, Pakistan) by using SVSV, MSVSV and conventional Fourier Transform methods.