The Research On Adaptive Symbol Recognition Algorithm Based On Unknown Signal Sample

Recent several decades, with the development of the techniques of communication network and data link, especially the development of kinds of complex grouping, agile,and wide-range communication techniques,processing of communicaiton detecting signal becomes more difficult,which has been confused by processing of the blind signal already. Many Communication Reconnaissance Receiver must perform the functions of signals'modulation types recognition,important modulation patameters estimation and so on,besides performing universal functions of accepting and processing.Automatic classification of modulation formats is the core of this thesis.Aiming at the kinds of the modulation formats and parameters of digital communication signals {MASK,MFSK,MPSK,16QAM}(M=2,4,8).firstly,identify modulation vast sorts using parameters based on transient characteristics is researched.secondly, under already known modulation vast sorts,identify modulation ranks each from using High-order cumulant and spectrum estimation to perform the auto-recognition algorithms.finally get the classification of detailed type of modulation siganals. Considering the signal types of practical communication system, symbol rate parameters estimation based on the feature of the Wavelet Transform (WT) is applied for the classification of analog and digital signals.Simulation demonstrates the method is feasible.In the process of parameters estimation,methods of signals' pre-processing based on spectrum analysis are studied,and parameters estimation of signal's carrier frequency,bandwidth,SNR and so on is performed.For digital communication signal,method of spectrum analysis based on the High-order cyclic cumulants for efficient carrier frequency estimation is put forward.