Sample clock offset must be estimated in order to correctly demodulate an OFDM signal. Estimates are also required in passive radar processing, to produce a reference signal, which is properly matched to the transmission. A recently proposed clock offset estimation method for passive radar is found to be overly sensitive to measurement noise. However, the radar ambiguity surface is found to be unaffected by the specific errors engendered by this noise. The reasons for this are explored. An updated method based on state-space modeling and Kalman smoothing is proposed. This is found to restore clock offset estimation performance in cases of higher noise, achieving this with low processing latency. These improvements enable the method to be employed for other applications requiring high quality, time-critical clock offset estimation.

中文翻译：

---

一种用于无源雷达的采样时钟偏移估计方法的改进

为了正确解调 OFDM 信号，必须估计采样时钟偏移。在无源雷达处理中也需要进行估计，以产生与传输正确匹配的参考信号。最近提出的无源雷达时钟偏移估计方法被发现对测量噪声过于敏感。然而，发现雷达模糊面不受该噪声产生的特定误差的影响。对此的原因进行了探讨。提出了一种基于状态空间建模和卡尔曼平滑的更新方法。发现这可以在噪声较高的情况下恢复时钟偏移估计性能，从而以低处理延迟实现这一点。这些改进使该方法能够用于需要高质量、时间关键的时钟偏移估计的其他应用。