Blind Estimation For Parameters Of Frequency-Hopping Signal Based On Time-Frequency Analysis

Frequency-hopping communication has been widely used in the military field when it is appeared because of its good anti-jamming, low probability of intercept and networking capabilities. The multi-path interference,the adjacent interference and man-made rifle-type interference can be effectively overcome by use of frequency hopping in FH communications technology and it has played a huge power in electronic warfare environment and improving communication anti-jamming capability of the most effective measures in the war.The three hot spots in modern signal processing-spectral estimation, higher order statistics methods, time-frequency analysis theory and technology are improving daily, and gradually be applied to the communication to combat areas, making the present communication countermeasure signal processing techniques, namely frequency-hopping signals surveillance and income, processing algorithm has reached a new climax.The frequency-hopping signal in frequency-hopping communication system is a typical time-varying, non-stationary signal. The time-frequency analysis technique of signal processing is a powerful tool when it is used to analyze non-stationary signals. Therefore, the main research of the time-frequency analysis technology is employed for blind frequency-hopping signal parameter estimation of this thesis.The thesis first simulates the Cohen class WVD, PWVD, SPWVD and Morlet wavelet transform by MATLAB,The corresponding time-frequency distribution map are obtained, and then the estimation for jump speed of frequency-hopping signal is also obtained. Secondly, in order to overcome the short-time Fourier transform due to the length of the fixed window function effects of time-frequency resolution to study, the long-adaptive window short-time Fourier transform which is time-frequency clusters of relatively high time-frequency diagram is studied. And a new algorithm is proposed by extracting the time-frequency ridge line, combined with a good Morlet wavelet's edge detection to estimate the jump speed of frequency-hopping signal.The research results shown that the algorithm could be overcome the defect of above algorithm which have lower estimate of the accuracy of the frequency-hopping signal's jump speed when the signal to noise ration is less than OdB.Finally, the results which through many simulation experiments under different SNR shown that the new algorithm can accurately estimated the jump speed of frequency-hopping signal when SNR is less than OdB but greater than-5dB.