Research On Azimuth Supper-resolution Algorithm For HF Surface Wave Radar

Compared to the conventional radar, High-Frequency Surface Wave Radar(HFSWR) own the unique advantages of distant detection range, providing a wide range of earlier vigilance time, and the warning time far beyond the normal radar, so it has high defense, political, military and economic value. However, the length of receiving array aperture is extremely limited by the actual size of the seaside antenna position, resulting in a poor performance of the angular resolution in the HFWSR. This paper mainly focuses on the HFSWR azimuth super-resolution algorithm research in the ideal array state and the non-ideal array state. The main works are as follows:1. Analysis the working principle of the HFWSR received signal theoretically, leads to the extraction process of DOA information. Carry out for the performance analysis and comparison of several existing classical super-resolution algorithm with a broader application, at the same time, validate the performance of MUSIC algorithm with HFSWR measured data.2. Aiming at the problem that angular resolution is limited to the array aperture in the HFSWR, and the traditional methods are difficult to break the rayleigh limit, a new model-AR cycle extrapolation, is proposed to solve this problem in this thesis. The method can achieve super-resolution by using the improved AR cycle extrapolation model to increase the length of the receiving data, equivalent to a virtual increase of receiving array aperture. Compared to the traditional super-resolution algorithm, the model adopted in this paper can extrapolate the received data reasonably under the single snapshot, so that not only the accuracy of estimated angle but also the azimuth resolution can be improved. Computer simulations and radar experimental data simulation s demonstrate the validity of the proposed model.3. Super-resolution algorithm discussed above are achieved under the assumption of ideal array manifold, namely, the array errors do not exist, and array steering vector precisely known. However, in practice, the array manifold errors are inevitable, hence it is necessary to analysis DOA estimation performance under non-ideal array state. The model of array response error, also an theory and simulation analysis of the error impact are given in this thesis. Compare the performance of three typical correction algorithm for array response error, the HFSWR systems often adopts self-correction technology base on various reasons. A new combination algorithm is presented in this thesis aiming at the traditional self-correction technology defects. The algorithm proposed adopts a combination of the passive correction algorithm based on covariance matrix with the iteration method, uses the estimated value of the former as an initial value of self-correction, and then iterate until convergence. Compensate the correction of the amplitude and phase error of the array by iteration method after pretreatment with initial value, then reuse the improved AR-MUSIC algorithm to improve the azimuth resolution of HFSWR. The computer simulation verified the effectiveness of the method.