A Fast Gridless Method for High Order Motion Parameter Estimation Based on PUMA

Abstract

This paper considers the efficient estimation of the high order motion parameters of the maneuvering targets. A fast gridless long-time coherent integration algorithm based on maximum likelihood (ML) estimation is proposed. After some matrix manipulations, we transform the ML estimation to a weighted least square (WLS) problem and solve it by Principal-singular-vector Utilization for Modal Analysis (PUMA). The proposed method avoids the grid effect with low computational complexity because the motion parameters are estimated by solving a polynomial equation instead of multi-dimensional searching. Simulations are conducted for validating the effectiveness and illustrating the high performance of the proposed method.