Design And Implementation Of FM Broadcast Monitoring System Based On Time Series Analysis

Radio spectrum resources are limited natural resources.In the frequency modulation broadcasting band,"black broadcasting" illegally occupies the scarce radio spectrum,which will not only interfere with the authorized broadcasting system,but also affect the normal operation of other communication systems such as aviation and emergency services.At present,the monitoring work of FM radio station mainly focuses on acquiring all occupied channels by collecting FM spectrum,and comparing these occupied channels with known station database,the broadcast frequency of unknown station can be obtained.At this time,the unknown radio stations are not all illegal broadcasting,so further judgment is needed.Most systems use a short broadcast recording to assist the judgment,which can achieve a certain monitoring effect.However,there is also a special form of illegal broadcasting.It is harmful to occupy the authorized broadcasting channel by using the closed rule of some authorized broadcasting during the day and at night in order to confuse the public,seriously affect the credibility of broadcasting institutions,directly defraud the audience's money and interfere with the broadcasting order.In this context,this paper proposes a FM broadcast monitoring system based on time series analysis.The core modules of the system are time series prediction module and spectrum classification module.The time series prediction module is to predict the time series of a broadcast channel,save the predicted value at the next time,and then compare the actual value with the predicted value at the next time.If the deviation between the two is large,it may cause abnormality,and then start the next frequency spectrum classification.The spectrum classification module uses a large number of authorized broadcast spectrum and unauthorized broadcast spectrum for training,and the training model is saved.The spectrum that may be abnormal in the previous step is tested with the model,and the classification results are obtained.This paper mainly explains the design and implementation of FM broadcast monitoring system based on time series analysis.The system collects FM spectrum through the spectrum receiver at the node end,normalizes the spectrum data,and detects the peak value.The number of channels and the corresponding center frequency in the range of 87 mhz to 108 MHz are obtained and stored in the database.Take the time series of a channel from the database for prediction.When the deviation between the actual result and the prediction result is large,the system recognizes that there is an exception,starts the spectrum classification work in the next stage,pushes the classification result to the cloud server,and then accesses the web page through the browser to obtain relevant information.