Design And Implementation Of FM Broadcasting Monitoring And Regional Localization System

Radio spectrum security is related to national network space security,and radio spectrum monitoring is the basis of radio spectrum management.In FM band,unlicensed broadcasting interferes with normal broadcasting systems and also disrupts the normal operation of other communication systems,especially those used in aviation and emergency services.In this paper,FM radio monitoring and regional location of unknown transmitters are studied,and the design and implementation of the monitoring system is carried out.The whole process of data acquisition,signal analysis,processing and monitoring results display is studied,and an automated and intelligent radio monitoring system for FM radio frequency band is realized.The main work of this paper is as follows:(1)A monitoring system architecture consisting of spectrum sensing node,server and data visualization layer is designed.Edge computing and software defined radio technology are used in the design of spectrum sensing node,which acquires the original monitoring data and send the monitoring results to the server.The server can further analyze,process and store the monitoring data,and the data visualization layer can display the monitoring information intuitively.(2)he discrimination between licensed and unlicensed broadcasting is studied.To detect FM broadcasting signal in spectrum data,the methods of threshold setting,clustering analysis and convolution neural network(CNN)are compared.The first two methods are susceptible to noise and DC spike.CNN based method converts spectrum into image for recognition,which can avoid the influence of noise to achieve accurate detection.A method for converting audio files into spectrograms and performing FM broadcast signal detection based on CNN classification is studied for audio data.After obtaining all the frequency points of the broadcasting signals,the discrimination of authorized and unlicensed FM broadcasters through comparing with the station database and frequency point combination classification is realized.(3)The regional location of the transmitter in the campus area was studied.Three spectrum sensing nodes are deployed in the campus,which is divided into 12 regions.A transmitter is located at each region to transmit broadcasting signal,while each node receive the spectrum data and send it to the server to construct a fingerprint database for further location of unknown transmitter.The regional location problem is turned to a classification problem,where machine learning methods are used to locate the transmitter.After the area is determined,the transmitter can be further located by collecting the received signal strength based on road test.The FM broadcasting monitoring function of the monitoring system has been used in the laboratory and Honghe State of Yunnan Province.The related technology has been applied for patent.