Abstract
An improved inverse filtering method is proposed which simultaneously estimates the coefficients of the recursive filter V(z) and the initial conditions for computing the output of V(z). By simulation, the resultant power spectral estimator with the initial conditions being estimated is demonstrated to have much smaller variance as well as a much smaller bias than that with the initial conditions wrongly set to be zero as the time constant of V(z) is comparable to the number of data. Results also indicate that some performance degradation to any spectral estimator involving the inverse filtering of data without estimating the initial conditions is inevitable as the number of data is not much larger than the time constant of the associated inverse filter.