Estimation of an Impulse Response Using Kurtosis
説明
In this paper, we propose a method of estimating an impulse response of a system by using kurtosis. The kurtosis is used for a measurement of the non-Gaussianity and the sharpness of probability distribution. The probability distribution of the sum of i.i.d. random variables is close to the Gaussian distribution in comparison with the original distribution. This theorem is called the central limit theorem. Based on the central limit theorem, we estimate the impulse response by maximizing the kurtosis of an estimated noise that is obtained from an observed signal and an input signal to the system.
