Scheduling Non-Preemptive Periodic Tasks in Soft Real-Time Systems using Fuzzy Inference

A Fuzzy-Based Multi-criteria Scheduler for Uniform Multiprocessor Real-Time Systems

“…Secondly, it works well with optimization and adaptive techniques. The third advantage of Sugeno type inference is that it is wellsuited to mathematical analysis [3]. Owing to the aforementioned benefits of Sugeno method, it has been applied in our proposed approach.…”

“…The satisfaction degree can then be used as a parameter for making a decision. One especially straightforward method to achieve this is the modeling of these parameters through fuzzy logic [3].…”

Untitled

“…To build a fuzzy system, inputs and output(s) to it must be first selected and partitioned into appropriate conceptual categories which actually represent a fuzzy set on a given input or output domain. Job priority, deadline and CPU time are selected as input to the Fuzzy Inference System (FIS) [53][54], which consist of five stages: The accuracy of schedulability analysis of system is largely depends on the accuracy with witch the worst-case execution time (WCET) have been derived which is nothing but the maximum amount of time each job can take to execute. The disadvantage of using WCETs under traditional task model like periodic, sporadic etc., is that system may be deemed un-schedulable even if they would function correctly most of the time when deployed.…”