Signal Modulation Recognition Algorithm Based on Optimal Neural Network

This paper presents a method of electromagnetic signal modulation recognition based on the combination of optimal neural network and characteristic parameter extraction. The method extracts four characteristic parameters and covers multiple information such as time domain, frequency domain and envelope of signal. At the same time, combined with the superior performance of the optimized neural network in nonlinear approximation, the recognition of modulated signals under low signal-to-noise ratio environment is realized. The simulation results show that this method has good anti-noise performance and recognition accuracy in signal modulation recognition.