Based On The FM Radio Band Adaptive Background Noise Extraction Algorithm Research

The background noise is one of the key factors in monitoring the impact of radiomonitoring in the procession of radio monitoring, and how to effectively extract the usefulsignal in noise is an important research direction. But many useful signal is submerged in a lotof background noise, which makes it difficult for us to detect.Nowadays, there are a lot of background noise extraction algorithms on radio, such as K-means clustering algorithm, the provisional value of the discriminant extraction algorithm.These algorithms can take apart of signal and noise under certain conditions, but each also haslimitations. At present, the problem can be broadly classified into two aspects. Firstly, thebackground noise of conventional extraction algorithms are able to achieve good results in theextraction of specific electromagnetic environment. However, with the change of the externalelectromagnetic environment, the background noise extraction algorithm as too dependent onthe supported hardware environments or set a fixed parameter can not be changed with theadaptation of the external environment, and thus appeared to detecting errors. Moreover,some background noise extraction algorithm can only be adapted to monitor certain frequencybands, which can not be applied to all monitoring requirements for monitoring band. As mostof the monitoring activities of the different frequency bands is different, the background noiseis also different, Similarly，the monitoring signal to noise ratio is not the same in the differentfrequency bands. Some monitoring bands even appeared more than one monitor business.This brings a degree of difficulty to the background noise extraction work.In this paper, features for the FM broadcast band. By segmentation method, smoothingfiltering and threshold selection method to start to build the algorithm. And thus presents anadaptive threshold segmentation algorithm based on smoothing, the algorithm uses both thesegmentation method and the smoothing algorithms to deal with general applicability to allmonitoring band. Meanwhile, the algorithm uses multi-frame signal processing methods tosolve the problem of changing the adaptability electromagnetic environment. Therefore, withthe change of external electromagnetic environment, the algorithm can be adaptively adjustedfetching background noise extraction, keeping the useful signal which can apply to mostmonitored frequencies, enabling the signal to noise separation.