Scheduling real-time systems with cyclic dependence using data criticality

Abstract

The increase of interdependent components in avionic and automotive software rises new challenges for real-time system integration. For instance, most scheduling and mapping techniques proposed in the literature rely on the availability of the system’s DAG representation. However, at the initial stage of system design, a dataflow graph (DFG) is generally used to represent the dependence between software components. Due to limited software knowledge, legacy components might not have fully-specified dependencies, leading to cycles in the DFG and making it difficult to determine the overall scheduling of the system as well as restrict access to DAG-based techniques. In this paper, we propose an approach that breaks cycles based on the assignment of a degree of importance and that with no inherent knowledge of the functional or temporal behaviour of the components. We define a “criticality” metric that quantifies the effect of removing edges on the system by tracking the propagation of error in the graph. The approach was reported to produce systems $$(56 \pm 14)\%$$ less critical than other methods. It was also validated on two case studies; a data modem and an industrial full-mission simulator, while ensuring the correctness of the system is maintained.