A Super-Resolution Multi-Target Location Algorithm for Frequency Modulated Continuous Wave Radar

To solve the problems of low resolution and high complexity of location algorithm of frequency modulated continuous wave radar, a super-resolution multi-target location algorithm is proposed by improving sparse theory, eigenvalue decomposition and fast Fourier transform. Firstly, we construct a linear antenna array signal model, and derive a general guidance vector suitable for both far and near fields. In order to reduce the complexity, the Taylor approximation of the antenna's two-way echo distance is obtained. Then, based on the sparse hypothesis, a small scale Fourier dictionary is constructed in the rough estimated Intermediate frequency. By establishing the regularization optimization model, the high precision estimation of the target Intermediate frequency is achieved which can realize the super-resolution location in the distance. Then, the Multiple Signal Classification is used to estimate the angle. The simplifyed of the guiding vector and the reducing of the data dimension can reduce the complexity. Based on the above model, multi-target localization is realized. Finally, the range resolution and spatial Angle resolution are calculated, the spatial sampling process is analyzed, and the superiority of the proposed algorithm in super-resolution and low-complexity is verified.