General orthonormal bases for robust identification in H?

Abstract

In this paper, the problem of system identification in H? with general orthonormal basis functions is investigated. A two-stage algorithm is shown to be robust, provided that the number of basis elements as a function of the amount of data does not increase faster than O(N2/5). Worst-case identification error bounds in the H? norm are derived. The algorithm also works on nonuniformly spaced frequency response measurements.