Robust time delay estimation based on asinh transform under α-stable noises

α-stable distribution is a sort of non-Gaussian random distribution and is widely applied to radar, sonar, atmosphere and biomedicine process field and so on. The salient feature of these random signals or noises are that the spike pulse is more significant than conventional Gaussian signal or noise. When dealing with non-Gaussian noises, the time delay algorithm using fractional lower order covariance (FLOC) can achieve better results, but it requests usually the previous estimation of characteristic exponent. In this paper, an adaptive time delay estimation algorithm using asinh transform is proposed. Compare with FLOC algorithm, the new aproach doesn't need the previous estimation of characteristic exponent, through changing the signal into a Gaussian with inverse hyperbolic sine transform under α-stable distributed noise condition, the new method can reliably converge and has got a good estimated precision.