Research On Automatic Modulation Recognition Of Digital Communication Signal

Modulation format is one of the most important characteristics used to distinguishcommunication signals. It can be used to monitor the activities of these signals, identify theircharacteristics, even to intercept the signal information content. With rapid development ofcommunication technology，the system and modulation manner of communication signalsbecome increasingly complicated and various. The modulation recognition of communicationsignals has important significance and widely applications in military and civil.Firstly, this paper analyzes the basic principle of common digital modulation andorthogonal transformation. Basing on the methods available of communication signals featureextraction, aiming at time domain and frequency domain properties of signals, an improvedmethod is proposed to recognize the signals including{2ASK、4ASK、2FSK、4FSK、2PSK、4PSK、16QAM}. And the simulation results indicate that the improved recognition algorithmis superior to the traditional method. Based on the previous researches, a method recognizingbase-band signals based on high-order cumulants is discussed in this paper. Simulation resultsshow that the accurate identification rate of every signal is higher than99%when SNR is2dB,so the method is still effective even when the SNR is very low. What’s more, depending onthe method of neural network, BP neural network is used to recognize the digital signalsincluding{2ASK、4ASK、2FSK、4FSK、2PSK、4PSK}, and achieve the anticipated effects.On the other hand, this paper also introduces the structural principle of TMS320F2812,and sets the digital signals modulation recognition platform basing on digital signal processor.Hardware emulation is done with CCS, and it indicates that the parametric statistical methodpresented in this dissertation is efficient.