Time Delay Estimation Algorithm Based On Correntropy Under Impulsive Noise Environment

As an important research content and component in the field of signal processing technology,time delay estimation has received great attention from experts and scholars at home and abroad in recent years.In previous research on time delay estimation,people usually assume that the noise in the receiving sensor conforms to the Gaussian distribution.However,the noise in the actual wireless communication and underwater acoustic communication environment has obvious spike impulse characteristics and their probability density distribution has a thicker tail than the Gaussian distribution.In this situation,the alpha-stable distribution noise model is more suitable to describe this kind of noise.Based on the basic theory of correntropy,this thesis puts the emphases on the robust single path time delay estimation algorithms and high-resolution multipath time delay estimation algorithms in a alpha-stable distributed impulse noise environment.Firstly,the problem of high-resolution multipath time delay estimation algorithms in a alpha-stable distributed impulse noise environment are studied.Aiming at the performance degradation of the expectation maximization(EM)multipath time delay estimation algorithm based on data addition and the weighted RELAXation(WRELAX)multipath time delay estimation algorithm based on the signal separation theory due to the impact of impulse noise,this thesis combines the maximum correntropy criterion(MCC)in the correntropy theory to improve the cost function based on the minimum mean square error criterion in the EM algorithm and the cost function based on the least square criterion in the WRELAX algorithm,and proposes the MCC-based Correntropy Expectation Maximization(CEM)algorithm and the MCC-based Correntropy WRELAX(CWR)algorithm respectively.The simulation experiment results show that the CEM algorithm and the CWR algorithm neither rely on the prior information of the impulse noise,but also can obtain better time delay estimation accuracy in a lower SNR environment,and the new algorithms show strong adaptability under noise with different impulse characteristics.Secondly,this thesis studies the time delay estimation problem of the two-sensor model in a alpha-stable distributed impulse noise environment.According to the property of the centered correntropy induced metric(CCIM),the property of the centered correntropy of the two received signals in the two-sensor model is analyzed.Based on this property,a two-sensor model delay estimation algorithm based on the maximum centered correntropy(CCETDE)is proposed,which effectively solves the problem that the peak sharpness of the corresponding time delay in the two-sensor model time delay estimation algorithm based on the correntropy(CETDE)is not obvious.Furthermore,the adaptive time delay estimation algorithm based on the maximum centered correntropy(MCCTDE)is studied.Simulation experiment results show that the MCCTDE algorithm can not only effectively suppress impulse noise,but also its estimation performance is better than the adaptive time delay estimation algorithm based on fractional lower-order statistics(LMPTDE)and the adaptive time delay estimation algorithm based on maximum correntropy(MCTDE).