FAST SEQUENTIAL ALGORITHMS FOR PARAMETER-ESTIMATION

This paper elucidates the parameter estimation basis common to adaptive filtering. A general formulation for developing two fast sequential algorithms is presented, capable of estimating adaptive processes accurately with a reduced number of parameters at the expense of extra processing: the time-domain parameter estimation algorithm; and the frequency-domain parameter estimation algorithm. Computer simulations for both low- and high-order processes are given to demonstrate the excellent performance of the proposed technique.