An Efficient Parameter Estimation Algorithm of the GTD Model Based on the MMP Algorithm

In this paper, a MMP algorithm is proposed. Compared with the classical algorithm, the proposed algorithm reduces the noise threshold of stably extractable parameters by 15 dB and reduces the computing time by ten or more. As an efficient way to interpret the measurements of high-frequency Inverse Synthetic Aperture Radar (ISAR), the Geometric Theory of Diffraction (GTD) model provides concise features of complex targets. However, the existing parameter extraction algorithms suffer from high computational complexity and poor ability against noise. To solve these challenges, a Maximum Matching Pursuit (MMP) algorithm is proposed in this paper. The proposed algorithm achieves parameter estimation by searching the maximum value of the dictionary matrix and observation signal matrix product. Compared to the classical OMP algorithm, the proposed algorithm significantly reduces the computational complexity by estimating params without cycles. To demonstrate the reconstruction efficiency of the improved algorithm, the Root-Mean-square-error (RMSE), and the computer time of the proposed algorithms are compared with the original algorithms, such as the OMP and the Estimating Signal Parameters via Rotational Invariance Techniques (ESPRIT) algorithm, under different Signal-to-Noise-Ratio (SNR). The simulation results show that the proposed algorithm reduces the noise threshold of stably extractable parameters by 15 dB, reducing the computing time by ten or more.