Research And Implementation Of Illegal FM Radio Signal Recognition System

After the state of the radio equipment market, The number of radio transmitting equipment is increasing. Many criminals used the FM radio transmitting equipment to occupy the normal frequency of all illegal activities, which even had affected the social stability, people’s lives and property safety. This made the radio management more complex, difficult. Illegal radio problems need to be addressed urgently. In this context, aiming at the above problems, the spirit could better serve the radio management, to investigate illegal FM radio signal recognition system.This system used the BP neural network algorithm for classification of FM broadcast signal, building up a three layer BP neural network, set its parameters through experiments. Through the monitoring center of Xihua University acquisition real-time data of legal signal radio FM radio frequency FM91.4MHz and illegal signal. By comparison of experiments to select five features of the representative values(mean, variance, standard deviation, maximum, median) as the input of the BP network. The output signal of illegal and legal signal for a sample of two expectations were labeled as 0 and 1, the network is trained, the formation of stable network. The training network will be written into the form of function by MATLAB, save the file as M, and as DLL.NET type component, easy call behind us write FM broadcast signal recognition system at.This system researched FM broadcasting recognition by Visual Studio 2008. First of all, we has designed four functional modules of the client interface, they are spectrum graphics real-time data display module, monitoring station and frequency information display module, the FM signal recognition result display module and function control module. System connected HuaRi monitoring equipment of Xihua University radio management center through RMTP protocol, realizes the communication and real-time data acquisition and display. The mixed programming of.NET and MATLAB, DLL component loading MATLAB formation and its reference, for processing real-time data received, calculate the characteristic value, and the characteristic value of the input to the BP neural network recognition, according to the monitoring results.Aiming at the problem of illegal radio station, we selected FM91.4MHz frequency test, designed FM broadcast signal recognition system based on BP neural network.