In the first two tracks controller design methodologies are derived for deterministic networks. Nevertheless, the resulting controllers are feedbacks which can also be applied when inter-arrival times and processing times are stochastic. First simulation experiments ([#References|references]) showed some promising results. In ([#References|references]) it was shown that the methodology results in distributed controllers as not yet considered in literature. It is worth to analyze the properties of controllers derived by the methodology obtained from Tracks 1 and 2 when inter-arrival times and processing times are stochastic. As already mentioned, this will be joint work with the Stochastic Operations Research Group of prof. Boxma at (Mathematics and Computer Science, TU/e) and EURANDOM.
Furthermore, when inter-arrival times and/or processing times are stochastic but unknown, and the resulting network controller can be parameterized, it is possible to determine the correct parameter setting online from measured data.