Research On Suppression Of The First Order Sea Clutter For Shipborne High Frequency Surface Wave Radar

The traditional microwave radar can not detect beyond the horizon. So it can not meet the demands of early warning and protection of exclusive economic zones, etc. High Frequency Surface Wave Radar utilizes the principle of vertically polarized electromagnetic wave diffracting along the air-water interface. Its detection range achieve up to hundreds or even thousands of miles. compared to shore-based HFSWR, shipboard HFSWR makes broadening of the sea clutter spectrum for the movement of the carrier platform. So the low-speed targets will be submerged and difficult to be detected. This paper will focus on the detection of the targets in sea clutter and the estimation of their DOAs.The spreading sea clutter has space-time coupling characteristics. The specific frequency corresponding to the specific direction of sea clutter. While it is difficult to distinguish between the target and clutter in time domain, we can process in the space domain. Spatial orthogonal weighting sets the the direction of the sea clutter to be null. Then targets detection and DOAs estimation will be achieved. However, the target and clutter within the same main lobe will have to split the main lobe. So oblique projection operator is introduced instead of the orthogonal weight vector. Ideal oblique projection method can suppress clutter whithout effection on the signals. But actually, obtaining the ideal oblique projection operator is difficult. It can be obtained by the decomposition of covariance matrix. Combinations of the oblique projection operator and the MUSIC algorithm can achieve super-resolution with supression of the sea clutter. From the perspective of matched filtering, we assume the projection of oblique projection operator to be scan vector and propose a new oblique projection algorithm based on steering vector. The new algorithm does not appear to split the main lobe. It can realize signal detection and clutter suppression at the same time. The peak of power spectrum can be used as the DOA estimation of the target signal. Wavelet transform performs well on signal detection and remote sensing of the sea with the ashore HFSWR. The paper also tries to transplant it to the shipborn HFSWR.Based on the range and velocity unit which is get by signal detection, this paper studies the super-resolution algorithm based on the single snapshot. IMP algorithm achieves precise estimation of DOAs by supressing the detected signals iteratively. The algorithm is not suitable to deal with a single snapshot data directly without preknowledgement of the number of signals or noise power. So we introduce the method of estimating covariance matrix in reduction dimension and beamforming algorithm to propose an improved IMP algorithm for single snapshot. The improved algorithm improves the probability of resolution without sacrificing the performance of estimation error and MSE, especially at low SNR. By processing the measured datas, this paper proves the effectiveness of the two methods.