Transient Interference Excision in Over-the-Horizon Radar by Robust Principal Component Analysis With a Structured Matrix

The performance of high-frequency skywave over-the-horizon radar is affected by transient interference. Traditional methods are used to suppress the transient interference in several steps. First, the position of the transient interference is detected in the time domain, and then the corrupted time samples are set to zeros. Then, the partial data are reconstructed by interpolation. However, the performance of the target detection would be decreased if the interference corrupts a significant portion of the received signal. In this letter, a novel approach based on robust principal component analysis (RPCA) is proposed to suppress the transient interference. Using a Hankel structured matrix constructed by the received signal in a given range cell, the transient interference excision is solved by RPCA, and a modified singular value thresholding algorithm is proposed to address the convex optimization problem. Without transient interference location and data interpolation, the proposed method can remove the transient interference, and it is also helpful for noise suppression. The experimental results show the effectiveness of the proposed method.