Time Frequency Difference Parameter Estimation Method For Target Location Based On Satellite Platform

In the modern electronic information warfare system,the electronic reconnaissance technology plays an irreplaceable role,and the passive location system based on the satellite platform,as the key part of the electronic reconnaissance,has a higher and higher status.However,the existing dual-satellite time-frequency difference joint passive location technology has some problems.The estimation accuracy of the time difference and frequency difference parameters has a direct effect on the positioning accuracy.However,the time-frequency offset estimation algorithm,which is widely used nowadays,can not meet the needs of further improving the positioning accuracy,and the computational complexity needs to be further reduced,which requires an estimation precision that can improve the time-frequency offset parameters,and reduce the complexity of the algorithm.However,the time-frequency offset caused by satellite motion needs to be further corrected,so that it is closer to the estimation of time-frequency offset parameters of non-time-varying signals.When we are in normal communication,we will be interfered by the enemy signal,and the most difficult thing to deal with is interference at the same time and at the same frequency.It is also necessary to study the separation technology of mixed communication signals and interference signals.After estimating the time-frequency offset parameters,the localization algorithm using the time-frequency offset parameters also needs to be optimized.In this paper,the joint passive location technology of time-frequency offset parameters is studied in detail,and it is found that the estimation accuracy of frequency offset parameters is not good by the second order cross-modulus algorithm.In the paper,we compare several existing joint estimation algorithms,such as second-order cross-correlation algorithm,expectation-maximization algorithm,fourth-order maximum likelihood algorithm,etc..It can be seen that fourth-order maximum likelihood algorithm and the second-order cross-correlation algorithm have the best performance in time-frequency offset parameters estimation.However,the order of the fourth-order maximum likelihood algorithm is too large and the operation complexity is too high.The second order cross-modulus algorithm has the lowest computational complexity.So we can use the second order cross-modulus algorithm for rough estimation,and then use the second order frequency domain cross-correlation algorithm for accurate estimation.The carrier frequency estimation method is introduced to determine the range of frequency offset in advance,and then the time-frequency offset joint search based on this range can greatly reduced the computational complexity.Then,the time-frequency difference estimation and time-variable correction of time-varying time-frequency difference signal are studied in this paper.The time-varying consistency and simulation authenticity of time-varying signal are ensured by using the raised cosine filter to smooth the time-varying signal.After segmenting the time-varying signals,the time-frequency offset parameters of each segment are estimated,and several time-frequency offset parameters are obtained,then all the estimated time-frequency differences of linear regression,find its rate of change.According to the rate of change,the estimation of time-frequency offset parameters can be corrected,and the result is close to the estimation of time-frequency offset parameters of non-time-varying signals.The performance of the non-sliding window segmentation method is compared with that of the sliding window segmentation method is compared with that of the sliding window segmentation method proposed in this paper.It can be seen that the sliding window segmentation method proposed in this paper can not only guarantee the estimation accuracy of the time-frequency difference parameters of a single segment signal,it can also increase the number of time-frequency difference estimation and increase the regression precision of linear regression.Then,the time-frequency difference parameters of the enemy jamming signal are estimated by separating the communication signal and the jamming signal.The second-order frequency-domain cross-correlation function is additive,that is,the second-order frequency-domain cross-correlation function of the mixed signal is subtracted from the second-order frequency-domain cross-correlation function of the known communication signal,the second-order frequency-domain cross-correlation function of the unknown enemy jamming signal can be obtained.The signal fading,time delay and frequency offset can be calculated according to the relative position of the communication signal source and the satellite,and then the time difference and frequency difference between the communication signal source and the two satellites can be calculated,the second-order frequency-domain cross-correlation function value of the communication signal varying with time and frequency corrections is calculated and the physical quantity is subtracted,the second-order frequency-domain cross-correlation function of the interference signal with time and frequency corrections can be obtained.Since the fading of communication signals is a statistical value rather than a constant value,the peak-to average ratio,that is,the cleanest case in which the correlation components of communication signals are removed,the time-frequency difference parameter of the interference signal can be estimated accurately.Finally,we use weighted least square algorithm,semi-definite programming algorithm and Newton iteration algorithm to deal with the time difference and frequency difference,and compare their advantages and disadvantages.In addition,some improved algorithms,and the positioning accuracy is improved under the condition of constant time-frequency difference estimation parameter error.