Probabilistic Response Time and Joint Analysis of Periodic Tasks
27th Euromicro Conference on Real-Time Systems
In this paper we address the problem of computing the probability response time distribution of periodic tasks scheduled on a uniprocessor system. Our framework assumes an arbitrary non-idling preemptive scheduling policy that may be either a fixed-priority scheduler (such as Rate Monotonic - RM) or a dynamic-priority scheduler (such as Earliest Deadline First - EDF). At the same time, our framework can handle arbitrary execution time distributions and arbitrary deadlines providing numerically accurate results. We also show how the framework can be extended to compute the correlation coefficients between the response times of different jobs by performing the joint analysis.
[TDEP15] Bogdan Tanasa, Unmesh D. Bordoloi, Petru Eles, Zebo Peng, "Probabilistic Response Time and Joint Analysis of Periodic Tasks", 27th Euromicro Conference on Real-Time Systems