BACKGROUND electromagnetic noise extraction Clustering Algorithm - based on hierarchical

The present invention discloses a layered based - electromagnetic background noise extraction clustering algorithm, relates to the field of radio management. Construction of the extraction method first spectral scans acquired spectrum monitoring - field strength data set; and the spectrum - field strength data sets and the data set into sub-field delamination strength values ​​acquired central field clusters each cluster; of the central field strength value clades sort, an electromagnetic background noise sample values; final sequence of samples forming the electromagnetic background noise, the background noise level acquisition electromagnetic unit time. The present invention improves the degree of automation and scientific measurement of electromagnetic background noise; effectively compensate for noise curve ITU-RP.372 method of analysis does not consider the actual radio system transmitter factors, thus easily resulting in insufficient noise estimate value representing the true value low; to overcome the ITU-R? SM.1753 rely on manual methods to select an idle mode frequency field sample values ​​as noise, improve the accuracy of calculation of electromagnetic background noise.