一种余弦调频信号增强新算法研究

A Special Algorithm for Enhancement of Cosine Frequency Modulation Signals

为了克服常规LMS算法抑制高斯噪声效果差、跟踪能力弱、且步长为零时无跟踪能力等缺点,利用高阶累积量抑制高斯噪声的性能,并将泄漏LMS算法和变步长LMS算法的优点统一在高阶累积量递推估计算法中,提出了基于高阶累积量泄漏-变步长类LMS算法迭代的余弦调频信号自适应增强新算法,对其原理进行了剖析。用实测运动目标辐射噪声数据,进行了动态仿真。结果表明:该算法能有效地克服基于LMS算法的自适应谱线增强器的不足,具有良好的抑制高斯有色噪声能力和跟踪性能。因此,本研究为水下运动目标识别与检测提供了新的技术途径。在工程实践中,具有重要的指导意义和应用价值。

When we use the conventional LMS-based ALE (adaptive line enhancement) algorithm to improve ability of underwater detection system to detect underwater target, there are three disadvantages: performance of handing Gaussian noise is bad, ability to trace time-varying signals is weak, and there will not be the ability to trace time-varying signals when the step size is zero. For greatly reducing these three disadvantages, we developed higher-order cumulant-based leakage-variable step size quasi-LMS adaptive line enhancement algorithm using the performance of cumulants in suppressing Gaussian noise and the advantages of leakage LMS algorithm and variable step size LMS algorithm, and analyzed its principle home. Based on the measured data radiated by the underwater moving-target, the enhancement process of cosine frequency modulation signals is dynamically simulated by the proposed algorithm. The simulation results show that the proposed algorithm can effectively overcome the shortcomings of the traditional LMS-based ALE algorithm and has a great capability of suppressing Gaussian colored noise and the performance of tracing time-varying signals. Accordingly, the research, in the paper, provides a new technical means for the recognition and detection of underwater moving-target.