基于矩阵奇异值分解的高频雷达瞬态干扰抑制

雷电等瞬态干扰严重影响了高频雷达的工作性能，必须加以抑制。该文提出了基于矩阵奇异值分解的高频雷达瞬态干扰抑制方法。该方法将高频雷达回波信号分段构造成矩阵并进行奇异值分解，首先根据矩阵有效秩的大小判断雷达回波中是否存在瞬态干扰，然后利用奇异值分解的正交性实现雷达回波的正交分解，使瞬态干扰分离出来，以利于检测，最后通过建立线性预测的全极点自回归模型对瞬态干扰位置处的回波信号予以恢复。实测数据处理结果表明该方法是有效的。

A SVD-Based Approach of Suppressing Transient Interference in High-Frequency Radar

The transient interferences, such as lightening impulse, meteor trail echoes and so on, have to be suppressed since they severely degrade the detection capability of High Frequency (HF) radar. An approach of suppressing the transient interference based on Singular Value Decomposition (SVD) is presented in this paper. In this approach, the HF radar echo is segmented and all segments constitute the columns of some matrix with its SVD computed. Firstly, it can be preliminarily judged whether the transient interference exists in the HF radar echo from the valid rank of the built matrix. Secondly, the orthogonal decomposition of the HF radar echo is obtained using the orthogonality of SVD, thus the transient interference is separated from the radar echo and so easily detected. Finally, the radar echo segment polluted by the transient interference is estimated by building the all-pole autoregressive linear prediction model. The effectiveness of this approach is demonstrated by examples of measured HF radar echo.