On metric dimensions of discrete-time systems (Q1199852)

In identification, information about an uncertain system in some universal space can be described by a subset that contains the uncertainty set of the system. This set is infinite dimensional and complicated in shape. For this reason the uncertainty is approximated by \(\varepsilon\)-nets (in the Kolmogorov sense). The paper deals with the description of these nets for discrete system of the form: \[ (Ku)_ t= \sum_{-\infty}^ \infty K_{t-k}u_ k \] where \(u\in \ell^ \infty\), \(K_ t=0\), \(t < 0\).