基于三阶累积量的水声信号时延估计研究

摘要

针对海洋探测接收信号信噪比低导致水声信号时延估计困难的问题, 提出了将三阶累积量的一维切片作为高阶统计量进行时延估计的新方法.该方法通过希尔伯特变换, 对相关峰进行锐化处理, 提高了时延估计精度.与直接相关法相比, 此方法不仅可以在较低信噪比下获得良好的时延估计效果, 而且可以抑制相关背景噪声和非高斯噪声干扰的影响, 计算机仿真和水池实验验证了此方法在较低信噪比和背景噪声为相关噪声及非高斯色噪声情况下的有效性.此方法为海洋探测中微弱回波信号的时延估计提供了一种新的技术途径.%In order to solve the problem of time delay estimation caused by low signal-to-noise ratio (SNR) of received signals in ocean exploration, a new method of time delay estimation is proposed by using one-dimensional slices of third-order cumulants as high-order statistics. This method improves the accuracy of time delay estimation by combining Hilbert transform to sharpen the correlation peaks and compared with direct correlation method. The proposed method can not only obtain good time delay estimation effect at low SNR, but also restrain the influence of correlated background noise and non-Gaussian noise. Computer simulation and water pool experiments verify the effectiveness of the proposed method with low signal-to-noise ratio and correlated background noise. This method provides a new idea of time delay estimation for detecting weak echo signals in ocean detection.