Work-in-Progress: Scheduling of Graph-Based End-to-End Tasks for Distributed Multi-criticality Systems
Juanjuan Wang, Hongan Wang · 2017
In this paper, we propose a multi-criticality graph-based end-to-end (MCE2E) task model. This task model is abstracted from the fault diagnosing and fixing process of the industrial control systems. The task in this model consists of a collection of nodes representing mixed criticality modes. Each node is defined by a parallel directed acyclic graph, the subtasks of which are pre-allocated to multiprocessors. We propose a preliminary scheme of the multi-criticality scheduling heuristics including the intra-criticality-mode graph structure for tasks at initial criticality modes, as well as the inter-criticality-mode changes via cluster grouping with the node adaption and time partitions for recovery and isolation.