A Temporal Approach to Stochastic Network Calculus

Stochastic network calculus is a newly developed theory for stochastic ser-vice guarantee analysis of computer networks. In the current stochastic net-work calculus literature, its fundamental models are based on the cumulativeamount of traffic or cumulative amount of service. However, there are net-work scenarios where direct application of such models is difficult. This paperpresents a temporal approach to stochastic network calculus. The key idea isto develop models and derive results from the time perspective. Particularly,we define traffic models and service models based on the cumulative packetinter-arrival time and the cumulative packet service time, respectively. Rela-tions among these models as well as with the existing models in the literatureare established. In addition, we prove the basic properties of the proposedmodels, such as delay bound and backlog bound, output characterization,concatenation property and superposition property. These results form atemporal stochastic network calculus and compliment the existing results.