FIR system identification using a linear combination of cumulants

A general linear approach to identifying the parameters of a moving average (MA) model from the statistics of the output is developed. It is shown that, under some constraints, the impulse response of the system can be expressed as a linear combination of cumulant slices. This result is then used to obtain a new well-conditioned linear method to estimate the MA parameters of a nonGaussian process. The proposed approach does not require a previous estimation of the filter order. Simulation results show improvement in performance with respect to existing methods