文中分析了声信号中随机噪声去除的可能性，讨论在时域上滤除随机噪声的算法。在分析和比较了固定滤波、自适应滤波等在时域上滤除噪声方法的基础上，提供使用最小均方自适应滤波的方法，研究了最佳滤波因子的获得和滤波参数的选择。用很小的运算量取得了较好的效果。 This paper analyzes the possibility of random noise removal in acoustic signals and discusses the algorithm of filtering out random noise in the time domain. Based on the analysis and comparison of the methods of filtering noise in the time domain, such as fixed filtering and adaptive filtering, the method of using least mean square adaptive filtering is provided and the selection of optimal filtering factor and filtering parameters are studied. With a small amount of computing has achieved good results.