Study On Lfm And Digital Signals Processing And Parameters Identification Based On The Time-Frequency Analysis

With the development of modern communications technology, non-cooperation communication play a more and more important part in comunications filed.Espectialy when software radio appears,the signal detection,modulation recognition and parameters estimation were gradually concerned by more and more people. Based on previous studies, this paper proposed a new algorithm about modulation recognition for digita signals and LFM signal, and then researched the detection and parameters estimation of LFM signal based on typical time-frequency analysis theories.As a typical non-stationary signal, LFM is widely used in comunication filed、radar systems and sonar and so on. So the LFM signal detection and parameters estimation are harder than stable signals.It’s urgent to find fast and effective algorithm with low computational complexity in future electronic war.First,this paper talk about the LFM signal model and the time-frequency analysis theories,The performances and difference of various time-frequency analysis methods.and then presents the definition of signal in time,frequency,time-frequency field and their relationship.The performances of various time-frequency analysis methods about the LFM signal detection plus noise are analyzed via experiments too.This paper also proposed an new parameter about modulation recognition based on characteristics of the signal’s power spectrum. Simulation results and mathematical operations shows the effective and the performance of the new parameters.When the detection and estimation of LFM signal is concerned, first,we discuss the method of Wigner-Ville distribution and ambiguity function. The performances of the two time-frequency analysis methods about the LFM signal detection plus noise are analyzed. Because of the linearity of LFM signal in time-frequency plane and the feature of detecting straight line of Radon transform in image processing,the methods of combining Radon transform with some typical time-frequency analysis methods,Such as WVD and Ambiguity Function,can well detect LFM signal embedded in noise.This paper also concerned in the process of LFM signal based on FRFT,including definition,properties and so on.The performances of WHT,RAT,FRFT are analyzed and compared both in computing complexity and in anti-noise performance, and the advantages and deficiencies are indicated based on experiments.