介绍了噪声抵消的原理和从强背景噪声中采用自适应滤波提取有用信号的方法:对基于Sigmind函数的变步长算法、基于箕舌线和基于抽样函数的变步长算法进行了对比研究。计算机仿真结果表明,这3种算法都能通过有效抑制各种干扰来提高强噪声背景中的信号检测特性;输出均方误差曲线和信噪比表明:基于抽样函数的变步长LMS算法具有良好的收敛性能、更小的权噪声和更大的抑噪能力。
The theory, of noise canceling and methods to abstract the desired signals from strong background noise using adaptive filtering are described. LMS algorithms based Sigmind function, Tongue-Like Curve ,and Sample function are compared. The simulation results show that all this algorithms can improve the ability to detect weak signals under the strong background noise. Compared with the other two algorithms, LMS algorithm based on Sample function has better performance, lower misadjustment noise, and stronger robustness against noise and disturbance.