Research On Modulation Analysis Of Digital Communication Signals In Alpha-stable Distribution Noise

Modulation analysis of digital communication signal plays an important role in the field of non-cooperative communication,which would be used in electronic reconnaissance,communication confrontation,spectrum monitoring and so on.Up to now,most of the communication signal modulation analysis methods assume that the ambient noise obeys the Gaussian distribution.In fact,there are significant short-time large-amplitude impulsive noises in nature,which is called impulsive noise and can be described by alpha-stable distribution.Since the stable distribution noise does not have the second order or the higher order statistics,many of the existing communication signal modulation analysis methods under the assumption of Gaussian noise are seriously degraded or even ineffective.In the literature,many blind algorithms have been proposed for MPSK signal modulation analysis in alpha-stable distributed noise,but lacks effective methods for other typical communication signals.This thesis deals with the problem of modulation analysis of typical digital modulation types with impulsive noise.Firstly,a blind processing algorithm based on generalized cyclostationary is proposed for OFDM signal modulation identification and parameter estimation in alpha-stable distributed noise.This scheme maped the received signal by nonlinear transformation,the analytical expressions for the generalized cyclic autocorrelation function are derived,and study the cyclostationarity aspect of nonlinear transformation SCLD and OFDM.The recognition statistics is structured based on generalized cyclic autocorrelation function to recognize OFDM versus SCLD signals.For the problem of blind parameter identification for OFDM,a novel algorithm based on generalized cyclic autocorrelation function is presented.Theoretical analysis and simulation results show that the validity of the novel algorithm in the Alpha stable distribution noise.Secondly,a modulation parameters estimation algorithm based on nonlinear transformation preprocessing is proposed for MFSK signal modulation order and symbol rate estimation in alphastable distributed noise.The nonlinear transform is used to suppress the noise impulse amplitude.The modulation frequency variation characteristic is extracted by the short-time Fourier transform for the nonlinear transformation MFSK signal,and the blind clustering algorithm is employed to estimate modulation order of MFSK signal.The symbol rate is estimated based on the wavelet transform of the nonlinear transformation MFSK signal.Simulation results show that the proposed algorithm has a good estimation performance for MFSK signal modulation order and symbol rate in stable distributed noise.Finally,a modulation analysis algorithm and processing is presented for the typical digital communication signals in apha-stable distributed noise.With the presented algorithm for signal modulation order and symbol rate estimation,a simulation system is designed for typical digital communication signals modulation analysis.Experimental test show that the extended algorithm and processing is feasible.