Carl von Ossietzky Universität Oldenburg
Unlimited capacity parallel processing models possessing different stopping rules are presented. These stopping rules imply inequalities on the modelsī response time distribution functions that can be tested empirically. This generalizes previous results on horse race models (minimum stopping rule) and exhaustive models (maximum stopping rule). Several areas of application are discussed and the results are demonstrated on a numerical example with conditionally independent exponential channel processing times.
This paper presupposes elementary probability concepts (some ofwhich are provided in the appendix). Some combinatorics and realanalysis are used for the numerical example in Section 4.