Double-Loop Feedback-Based Scheduling Approach for Distributed Real-Time Systems.pdf

下载文档

11
0
约3.14万字
约 10页
2017-04-11 发布于江苏
举报
版权申诉
保障服务

Double-Loop Feedback-Based Scheduling Approach for Distributed Real-Time Systems.pdf

1、本文档共10页，可阅读全部内容。
2、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

Double-Loop Feedback-Based Scheduling Approach for Distributed Real-Time Systems

Double-loop Feedback-based Scheduling Approach for Distributed Real-Time Systems Suzhen Lin and G. Manimaran Department of Electrical and Computer Engineering Iowa State University, Ames, IA 50011, USA linsz,gmani @iastate.edu Abstract. The use of feedback control techniques has been gaining importance in real-time scheduling as a means to provide predictable performance in the face of uncertain workload. In this paper, we propose and analyze a feedback schedul- ing algorithm, called double-loop feedback scheduler, for distributed real-time systems, whose objective is to keep the deadline miss ratio near the desired value and achieve high CPU utilization. This objective is achieved by an integrated de- sign of a local and a global feedback scheduler. We provide the stability analysis of the double-loop system. We also carry out extensive simulation studies to eval- uate the performance and stability of the proposed double-loop scheduler. Our studies show that the proposed scheduler achieves high CPU utilization with low miss ratio and stays in steady state after a step change in workload, characterized by change in actual execution time of tasks. 1 Introduction and Motivation In real-time systems, the correctness of the system depends not only on the logical cor- rectness of the result, but also on the time at which the results are produced [1][2]. Tra- ditional real-time scheduling algorithms are based on estimations of the (pessimistic) worst-case execution time ( ) of tasks. However, this will result in an extremely underutilized system. In many cases, it is preferable to base scheduling decisions on a more correctly estimated execution time ( ) and to be ready to deal with changes in execution times dynamically. This approach is especially preferable as it provides a firm performance guarantee in terms of deadline misses while achieving high throughput. One of the very successful areas in addressing performance in the presence of un- certainty