Research On TDOA Parameter Estimation Based On Three-satellite Interference Location

With the wide use of communication satellites, satellite incidents of interference areincreasing and the threats of satellite communication system are also increasing.According to the report, it appears hundreds of satellite interference events every year,and these events are showing the trend of growth in the number. Satellite interferencelocation has become a new research topic of current security of satellitecommunication system. The process of satellite interference location has two steps:first, estimate the corresponding parameters; then, using some algorithms and theestimated parameters to compute the location of the interference. Three-satelliteinterference location is based on two TDOA parameters, which are estimated by theinterfered satellite and two adjacent satellites. We can get the area of interferencesource using of the intersection between two TDOA lines and the Earth’s surface, soas to realize the location of interference sources. Therefore, method of parameterestimation is the key to realize positioning problem.In this paper, we mainly study the methods of three-satellite TDOA location in thefield of passive location. Combined with fast Fourier transform, wavelet transform，Hilbert transform, the high order cumulants and adaptive filtering and so oncorrelation theory, this paper have made in-depth research for the second method,wavelet transform correlation method and fourth-order cumulant-adaptive TDOAestimation algorithm.The main work and research achievements of this paper can be concluded asfollows:(1) To solve the real-time requirements of satellite interference location, this paperproposes the fast secondary TDOA estimation algorithm. The algorithm based onsecond correlation TDOA estimation uses the characteristics of the Fourier transformfor correlation function and power spectral density to each other; using the FFT (FastFourier Transform) for rapid calculation to reduce the amount of calculation, canquickly and accurately estimate TDOA parameters of signal.The simulation andexperimental results show that in the same SNR, the proposed algorithm can quicklyestimate TDOA, and effectively improve the real-time requirements of satelliteinterference location.(2) To solve the higher precision issues of TDOA estimation, based on the research of wavelet transform de-noising and Hilbert transform, this paper proposescorrelation and Hilbert subtraction TDOA estimation algorithm based on waveletdenoising. This algorithm utilized the characteristics of the zero-crossing point ofHilbert transform method corresponding to the correlation peak point of correlationmethod, using correlation function of signals after wavelet transform minus theabsolute value of its Hilbert transform, to sharpen peak point and get high TDOAestimation precision in the relatively low SNR, regardless of the signal is stationary ornon-stationary.(3) To solve the problems of correlated noise and variable signal-to-noise ratio,based on the study of higher-order cumulant and Least Mean Square(LMS) adaptiveTDOA estimation algorithm, this paper proposes the improved Least Mean Squaretime delay estimation algorithm based on fourth-order cumulants(improvedFOC-LMSTDE). The focus of the algorithm is that the LMSTDE adaptive filter isdivided into two parts, so that the adaptive process of time delay and SNR separate:one is to track time delay, and the other is to adjust to changes of SNR. Coupled withthe advantage of the higher order cumulants can handle correlated noise, the proposedalgorithm not only can get valid TDOA estimation under the condition of correlatednoise and low SNR, but make TDOA estimation in variable SNR.