Research On Parameter Estimation Method Of Frequency Hopping Signals Based On Compressed Sensing

Frequency hopping communication are widely used in the field of communication because of its good anti-interference and confidentiality,especially in the field of military communications,and the frequency hopping signal monitoring technology has always been a hot issue in frequency hopping communication research.At present,frequency hopping communication is gradually developing towards high-speed and high-bandwidth.The traditional Nyquist sampling method is used,the data amount of the wideband frequency hopping signal will be huge,and the traditional processing method is complicated to calculate,resulting in the processing of the interception technique of the frequency signal becomes particularly difficult.Compressed sensing theory can process data in undersampling mode,combining compressed sensing theory with wideband frequency hopping signals can effectively solve the problem of wideband frequency hopping signal parameter estimation.In this thesis,the parameters of frequency hopping signal estimation based on compressed sensing are studied under different frequency hopping signal models.The main contents of the work include:(1)The existing parameter estimation method does not fully consider the frequency domain structure characteristics of the frequency hopping signal.Under the priori condition with known frequency hopping period,there is no need to reconstruct the frequency hopping signal.Firstly,the frequency is quickly estimated by using the segmentation compression.The frequency value is corrected and processed.Then,according to the modified frequency map,the existing parameter estimation method of hopping time is improved,and the parameters of the frequency hopping signal are effectively estimated while reducing the computational complexity.(2)Aiming at the problem that the traditional time-frequency analysis method has high computational complexity and low parameter estimation accuracy under low SNR conditions,it is improved on the basis of parameter estimation algorithm under the condition of prior information,a blind parameters estimation method based on compressed sending for frequency hopping signal.Combined with the idea of short-time Fourier transform,compress the signal in the window by compressed sensing,and the transition time and the skip period of the signal are estimated according to the distribution of the maximum inner product value.Since the estimated hopping time accuracy is not high at this time,by constructing a Fourier orthogonal base diagonal matrix,the traversal step size is improved,and the hopping time isaccurately estimated.The simulation results show that the proposed method can ensure the accuracy of parameter estimation at low SNR and reduce the computation time.(3)Most existing methods for estimating frequency hopping signal parameters are based on conventional frequency hopping signal models,and less research on unconventional frequency hopping signals.This section mainly analyzes the model of a unconventional frequency hopping signals.Combining the theory of compressed sensing with unconventional frequency hopping signals,a parameter estimation method for unconventional frequency hopping signals based on compressed sensing is presented.Since the hopping time of the unconventional frequency hopping signal is not regular,for this problem,the frequency estimation value is aggregated,the threshold is set,and eliminating frequency estimations with large errors,so that the estimation accuracy is higher.This method effectively reduces the computation time of parameter estimation and guarantees the accuracy of parameter estimation at low SNR.