The Spectral Properties of First-Order Networks Having a Binary Random Parameter Variation

The spectral properties of first-order networks containing an element which varies randomly between two values are derived. The results obtained apply for stationary binary parameter variations where the lengths of successive time intervals between changes in the parameter are statistically independent. Explicit solutions are obtained when the binary parameter variation is 1) a Poisson process and 2) a periodic process. In the case of the binary Poisson parameter variation experimental confirmation of the results for one particular filter is given. The experimental equipment is briefly described.