| Popis: |
The scheduling problem answers the question whether a given set of input tasks is feasible. It has been studied since the 1970s and many important results have been revealed to the scientific community. This paper presents an implementation tool, PEARLS, i.e., Pliable EArliest Deadline First, Rate Monotonic, Least Laxity Schedulers, based on some of the most significant existing schedulability analytical feasibility conditions and schedulability algorithms. If the input tasks set is feasible, our tool can simulate the popular scheduling methods: rate monotonic (RM), earliest deadline first (EDF) and least laxity (LL). Moreover, the tool is designed to handle both preemptive as well as non preemptive tasks. Both periodic and sporadic tasks together with or without precedence constraints may also be considered in the input. Experimental results have been conducted for our tool implemented in the recent Java Development Kit (JDK) version 1.6 on a Pentium GHz system having 1 MB memory. We tested our tool on several real-time systems specifications, and the obtained experimental results have confirmed that our tool is efficient and useful. |
