Effective resource provision for real-time applications in static-priority scheduling networks

In this paper, we propose an effective resource provision technique, which can provide absolute delay guaranteed services for real-time application. Based on other work on connection- population-insensitive delay computing and worst-case delay analysis, we develop methods which consider the run time network workload uneven distribution factors to configure resources provided to real-time applications. With these new methods, utilization based admission control can also be used to provide scalable and efficient admission control. We assume the underlying network to use static scheduling. We concentrate on the simplest case that only one class of real-time application is services. We compare the results of our adaptive strategy with those of the uniform UBAC method and demonstrate the improvement in effective utilization of system resources.