Adaptive fuzzy control scheduling of hybrid real-time systems
Xiaoyong Huai, Yong Zou, Mingshu Li · 2003
To meet the ever-increasing requirement of multi-task uncertainly concurring, dynamic uncertainty of tasks and system loads in open hybrid real-time systems, the paper presents a QoS model based open adaptive hybrid soft realtime systems scheduling framework. To adapt to the uncertainties of open hybrid real-time systems, we introduce fast fuzzy control mechanism to control the scheduling errors of QoS objects, so as to statistically guarantee that the accepted tasks of any QoS object can complete before their deadline. Simulation results illustrate that the scheduling framework can adapt the dynamic uncertainty of system objects, loads and tasks, and statistically guarantee the multi-constraints. It is an intelligent adaptive scheduling framework that can be applied to hybrid soft real-time systems.