Spectrum Sensing And Parameter Estimation Of Electromagnetic Signal In Alpha Stable Distribution Noise

Alpha stable distribution is similar to the characteristics of noise in the actual environment because of its burst property.Therefore,the distribution model is often used to model noise in the analysis and processing of electromagnetic signals,so it is more widely used.Especially,in radar,sonar,communication and other fields,the spectrum sensing and parameter estimation of electromagnetic signal in alpha stable distribution noise have become a hot issue in signal processing.Aiming at the problem that the existing spectrum sensing and parameter estimation algorithms have low performance in this noise and cannot meet the practical application requirements,this paper discusses it from the following three aspects.Firstly,two multi antenna spectrum sensing methods are proposed to solve the problem that the performance of traditional spectrum sensing algorithms is degraded in Alpha stable distributed noise,That is,spectrum sensing algorithm based on weighted diagonal elements of fractional low order covariance matrix and spectrum sensing algorithm based on weighted non diagonal elements of fractional low order covariance matrix.The two methods construct the detection statistics of multi antenna received signal vector by weighting the elements in the fractional lower order covariance matrix.The simulation results show that the two algorithms have better spectrum sensing performance than the existing sensing algorithms.Secondly,an algorithm based on Generalized Extended Linear Chirplet Transform(GELCT)and an algorithm based on Generalized Fourier Transform(GFT)are proposed to estimate the modulation parameters of LFM signal,The GELCT algorithm first uses a nonlinear transformation to obtain the LFM signal which suppresses the Alpha stable noise,and then carries on the time-frequency transformation to the signal.Then,Radon transform is carried out,and the maximum value after the transform is calculated.According to the corresponding angle position,the estimated value of chirp rate is obtained.The demodulated reference signal is constructed by using the estimated chirp rate,and the value obtained by multiplying the demodulated reference signal with the original received signal is processed by Generalized Fourier transform.The estimated value of the initial frequency of LFM signal is obtained by using the position of the maximum value after generalized Fourier transform.The simulation results show that the performance of the proposed method is close to its corresponding Cramér-Rao lower bound in Alpha stable distribution noise,and the performance of the proposed method is better than the existing algorithms in low SNR,with good estimation accuracy.Thirdly,the modulation frequency interval and symbol period of MSK signal are estimated.The generalized second-order cyclic statistics of MSK signal is calculated first,that is,the received MSK signal is transformed nonlinearly to suppress the alpha stable distribution noise in the received signal,and then the second-order cyclic statistics is calculated.Then,the specific delay cross section of generalized second-order cyclic statistics is extracted,and the cyclic frequency set is obtained by adaptive double threshold detection,the modulation frequency interval is estimated by using the interval between adjacent cyclic frequencies in the cyclic frequency set.Finally,the symbol period is estimated according to the modulation index of MSK signal.The simulation results show that the performance of the proposed method is close to its corresponding Cramér-Rao lower bound in Alpha stable distribution noise,and the performance of the proposed MSK parameter estimation method is better than the existing algorithms,and has good estimation performance.