Adaptive Fixed Priority End-To-End Imprecise Scheduling In Distributed Real Time Systems

TL;DR

This paper introduces an adaptive scheduling framework for distributed real-time systems that employs imprecise computation to improve dependability and predictability during overload conditions.

Contribution

It extends existing scheduling frameworks with a new priority scheme and utilization adjustment algorithm to better handle transient overloads using imprecise computation techniques.

Findings

Framework improves schedulability during overloads

Enhanced predictability and dependability over existing methods

Effective utilization adjustment algorithm implemented

Abstract

In end-to-end distributed real time systems, a task may be executed sequentially on different processors. The end-toend task response time must not exceed the end-to-end task deadline to consider the task a schedulable task. In transient over load periods, deadlines may be missed or processors may saturate. The imprecise computation technique is a way to overcome the mentioned problems by trading off precision and timeliness. We developed an imprecise integrated framework for scheduling fixed priority end-to-end tasks in distributed real time systems by extending an existing integrated framework for the same problem. We devised a new priority assignment scheme called global mandatory relevance scheme to meet the concept of imprecise computation. We devised an algorithm for processor utilization adjustment, this algorithm decreases the processor load when the processor utilization is greater than one. Also we extended the schedulability analysis algorithms presented in the old framework to allow adaptive priority assignment and to meet imprecise computation concept. Simulation results showed that our new framework is more dependable and predictable than the existing framework over transient overload periods.