A new adaptive time delay estimation using fourth-order cumulants based on polarity iterative

In order to shorten the computation time of high order statistics and improve the efficiency of LMS algorithms in suppressing the Gaussian noise, here we proposed a new fast algorithm in which the different arrival time of the linear non-Gaussian signal between two spatially separated sensors are estimated. It is assumed that the signals are corrupted by the unknown cross correlated Gaussian noises. With a polarity iterative algorithm, the computational complexity of cumulant is reduced. The simulation results show that the method works well under low SNR condition.