Graph-based models for real-time workload: a survey

This paper provides a survey on task models to characterize real-time workloads at different levels of abstraction for the design and analysis of real-time systems. It covers the classic periodic and sporadic models by Liu and Layland et al., their extensions to describe recurring and branching structures as well as general graph- and automata-based models to allow modeling of complex structures such as mode switches, local loops and also global timing constraints. The focus is on the precise semantics of the various models and on the solutions and complexity results of the respective feasibilty and schedulability analysis problems for preemptable uniprocessors.