Research On Detection And Recogniton Technology Of MSK Signal Based On Duffing Oscillators

In the process of information transmission and information confrontation,it is always important that signal detection and modulation recognition.In order to improve the power efficiency and security of communication signal,modern modulation techniques are widely used such as MSK,GMSK,spread spectrum,and so on.Because the power spectrum density of this kind of communication signal is low,it is difficult to detect and identify the communication signal which is submerged in strong noise by traditional information countermeasure methods.As a branch of nonlinear science,chaotic theory has the characteristics of immunity to noise and sensitivity to weak signals.It has shown great superiority in the field of weak signal detection.Accordingly,Duffing oscillator is attempted for MSK signal detection and modulation recognition in this dissertation.The characteristics of MSK are utilized that modulation signal phase is continuous,modulation frequency is orthogonal and frequency difference is minimum.Firstly,a model consist of double Duffing oscillators is established for MSK signal detection.And then,power spectrum entropy is used to determine the state quantitatively.And the process of signal detection is designed.Finally,an array system consist of four Duffing oscillators is designed.Power spectral entropy is used as the characteristic parameter.According to the principle that different modulate signals will cause different systems change.RBF neural network is designed as a classifier to classify and recognize.Therefore,2ASK,BFSK,BPSK and MSK signals are recognized.The simulation results verify the feasibility of the methods of detection and identification.Compared with traditional methods,there are many advantages that it has high detection sensitivity and high recognition accuracy for the technical methods which are mentioned in this dissertation.What is more,it is immune to noise and independent of prior information.In the condition of SNR as low as-20 dB,the system can still maintains stable and it has great robustness.