Robust Time Delay Estimation Method Based on Fractional Lower Order Cyclic Statistics

In the presence of additive impulsive noise, the cyclic correlation and other second cyclic statistics degenerate and the cyclostationary information about desired signals is lost. To combat this effect this paper proposes a novel cyclic correlation which is well-behaved in the presence of alpha stable noise. Based on the proposed definition a new adaptive time delay estimation method based on fractional lower order cyclic statistics is developed. Simulations show that the performance of the proposed algorithm is robust both in Gaussian noise environment and alpha stable distributed noise environment compared to the traditional second cyclic statistics based time delay estimation method and another typical LMP method.