中文摘要：

针对高频地波雷达中出现的瞬态干扰问题，提出了基于S变换（Stockwell transforms）的瞬态干扰抑制方法．首先分析了两次FFT信号处理机制下瞬态干扰的时频特征；然后应用s变换对雷达信号进行时频分析，利用瞬态干扰在时频维扩展性不同的特点，实现瞬态干扰的检测；最后应用三层BP（Back Propagation）神经网络对海洋回波信号进行预测和恢复．该方法避免了海杂波抑制的步骤，具有计算量小、信号损失小的优点．现场实测数据分析表明，该方法能够有效抑制瞬态干扰，使回波多普勒谱信噪比提高达10dB，有效提高了雷达的抗干扰能力．

英文摘要：

A new method based on S tansforms is proposed to suppress instantaneous interference for high frequency surface wave radar （HFSWR） application. The characteristics of instantaneous interference in time and Doppler domain were discussed within the twice-FFT signal processing framework; then S transforms was applied to outline the characteristics of instantaneous inter- ference in time-frequency domain for interference detection and excision; finally, BP neural network was used to recover the sea e- cho signals. Experiment results show that the anti-interference ability is enhanced by this method, and the SNR of Doppler spectrum is improved for about 10dB.