AI-Assisted FM Broadcasting Monitoring and Localization

Artificial intelligence (AI) is widely used in the fields of images, voice, games, etc., but very few is reported in the field of radio monitoring. In order to achieve automatic and intelligent FM broadcasting monitoring, an AI-assisted FM broadcasting monitoring and localization system which consists spectrum sensing node (SSN) cluster and server is presented in this paper. In the system, we monitor the radio spectrum in FM band, and use convolutional neural network (CNN) to classify the spectrum data and realize the discrimination between licensed and unlicensed broadcasting. For localization of unknown transmitter, three SSNs are deployed in the campus of Yunnan University, which is divided into 12 regions. A transmitter is located at each region to transmit broadcasting signal, while each SSN receive the spectrum data and send it to the server to construct a fingerprint database for further location of unknown transmitter. Through collecting the spectrum datasets in the experiment, both the received signal strength (RSS) and the spectrum data fingerprints-based localization strategies are implemented based on AI algorithms.