Model based design of super schedulers managing catastrophic scenario in hard real time systems

摘要

The conventional design of real-time approaches depends heavily on the normalperformance of systems and it often becomes incapacitated in dealing withcatastrophic scenarios effectively. There are several investigations carriedout to effectively tackle large scale catastrophe of a plant and how real-timesystems must reorganize itself to respond optimally to changing scenarios toreduce catastrophe and aid human intervention. The study presented here is inthis direction and the model accommodates catastrophe generated tasks while ittries to minimize the total number of deadline miss, by dynamically schedulingthe unusual pattern of tasks. The problem is NP hard. We prove the methods foran optimal scheduling algorithm. We also derive a model to maintain thestability of the processes. Moreover, we study the problem of minimizing thenumber of processors required for scheduling with a set of periodic andsporadic hard real time tasks with primary/backup mechanism to achieve faulttolerance. EDF scheduling algorithms are used on each processor to managescenario changes. Finally we present a simulation of super scheduler withsmall, medium and large real time tasks pattern for catastrophe management.