Study On Multipath Time Delay Estimation Algorithms Under Alpha-Stable Noises

Time delay estimation is the research which refers to estimate and determine the time difference caused by different receivers and its related parameters by using signal processing theories and parameter estimation methods. Time delay estimation has important theoretical significance, and also has a very wide range of applications in the fields of sonar, radar, biological medicine and petroleum exploration. Those signals received through multipath propagation are formed by the source signals and those superimposed signals. Under multipath environment, the primary task is how to decompose superimposed signals for purpose of estimating time delays in multi-arrivals.Gauss signal model is easy to carry out theoretical analysis, so it has been the dominant position in the traditional signal processing field. However, many signals and noises in actual application are non-Gaussian, such as low-frequency atmospheric noises, underwater acoustic signals and many other noises produced by human, these signals and noises have shown a notable pulse characteristic. It will result in noticeable degradation of the algorithms, because the model doesn’t match well with these signals or noises. People usually use Alpha-stable distribution to describe those signals and noises, and because Alpha-stable distribution is generalized Gaussian distribution, which means Gaussian distribution is a special case of alpha-stable distribution. So Alpha-stable model can be used in more widely range.In this thesis, based on the basic properties of alpha-stable distribution, several classical time delay estimation methods based on the second-order statistics are studied. Drawing on existing multipath time delay estimation algorithms based on the fractional lower order moment theory, we put forward two multipath time delay estimation algorithms with high resolution under different background noises:the generalized EM algorithm and the generalized WR algorithm. It shows that these new algorithms are robust for both Gaussian and non-Gaussian noise environment through computer simulations and theoretical analysis. And the four multipath time delay algorithms have been simulated for the comparison under different signal-to-noise ratios; this further proved the good toughness of these algorithms in this paper.