一种基于邻值比较的信噪分离方法

A Method of Signal-to-noise Separation Based on Neighbor Value Comparison

为更高效地监测射电天文台站电磁环境,需针对宽带频谱序列研究高精度信噪分离方法,为进一步的信号识别和统计提供算法支撑。依托实测频谱数据样本,提出一种基于邻值比较的信噪分离方法。首先,采用邻值比较算法,提取宽带频谱的背景噪声,并对提取背景噪声进行滤波处理,结合标准差理论,确定宽带频谱的信噪分离阈值,实现宽带频谱信号和噪声的分离。其次,采用此方法对实测数据样本进行处理,并与现有信噪分离方法进行对比。结果表明,该方法通用性较好,信噪分离准确度高,能够应用于射电天文台站的电磁环境实时监测。

In order to monitor the electromagnetic environment of radio astronomical stations more efficiently,it is necessary to study high-precision signal-to-noise separation methods for broadband spectrum sequences to provide algorithm support for further signal identification and statistics.A signal-to-noise separation method based on neighbor comparison was proposed.First of all,based on the measured spectrum data samples,the neighbor comparison algorithm was used to extract the background noise of the broadband spectrum.Combined with the standard deviation theory,the signal-to-noise separation threshold was given to realize the signal and noise separation.Secondly,the proposed method was applied to the measured data,and the algorithm accuracy and operating time were compared with the available signal-to-noise separation methods.The results show that the method has good versatility and high accuracy for signal-to-noise separation,which can be used to the real-time electromagnetic environment monitoring in radio astronomical stations.