Scheduling Model in Global Real-Time High Performance Computing with Network Calculus

We first put forward abstractive model of global scheduling pattern from the global scheduling pattern. Under the framework of the abstractive pattern, we can use a rate-latency server into the QoS parameters, although it can not be the optimal scheduling algorithms in the special schedulers, but for the global scale. It can be easy to get the delay bounds of end-to-end in the uncertain surroundings using some definitions and theories of network calculus.