Performance evaluation of distributed real-time scheduling algorithms

P. M Perry Leung, Jaideep Srivastava, W. T. Tsai

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A performance evaluation of the two classes of real-time distributed CPU scheduling algorithms (sender-initiated and receiver-initiated) has been conducted, with emphasis on several factors, including network load variance, task laxity, communication costs, and mean arrival rate. Results show the pivotal role of task laxity, communication cost, and network load variance in determining the performance of each algorithm. It was observed that, when the algorithms are run on a fully connected network, receiver-initiated algorithms outperform sender-initiated algorithms with low-laxity tasks. If the laxity of the tasks is high, then both sender-initiated and receiver-initiated perform well, and the difference is not significant. Due to the low communication overhead on a fully connected network, the receiver initiated algorithm can take advantage of a frequent polling scheme to maintain up-to-date peers' CPU surplus information. Consequently, it achieves a relatively low overall miss ratio.

Original languageEnglish
Title of host publicationProc First Int Conf Syst Integr ICSI 90
Place of PublicationPiscataway, NJ, United States
PublisherPubl by IEEE
Pages752-761
Number of pages10
ISBN (Print)0818690275
Publication statusPublished - 1990
Externally publishedYes
EventProceedings of the First International Conference on Systems Integration - ICSI '90 - Morristown, NJ, USA
Duration: 23 Apr 199026 Apr 1990

Other

OtherProceedings of the First International Conference on Systems Integration - ICSI '90
CityMorristown, NJ, USA
Period23/4/9026/4/90

Fingerprint

Scheduling algorithms
Program processors
Communication
Costs

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Leung, P. M. P., Srivastava, J., & Tsai, W. T. (1990). Performance evaluation of distributed real-time scheduling algorithms. In Proc First Int Conf Syst Integr ICSI 90 (pp. 752-761). Piscataway, NJ, United States: Publ by IEEE.

Performance evaluation of distributed real-time scheduling algorithms. / Leung, P. M Perry; Srivastava, Jaideep; Tsai, W. T.

Proc First Int Conf Syst Integr ICSI 90. Piscataway, NJ, United States : Publ by IEEE, 1990. p. 752-761.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Leung, PMP, Srivastava, J & Tsai, WT 1990, Performance evaluation of distributed real-time scheduling algorithms. in Proc First Int Conf Syst Integr ICSI 90. Publ by IEEE, Piscataway, NJ, United States, pp. 752-761, Proceedings of the First International Conference on Systems Integration - ICSI '90, Morristown, NJ, USA, 23/4/90.
Leung PMP, Srivastava J, Tsai WT. Performance evaluation of distributed real-time scheduling algorithms. In Proc First Int Conf Syst Integr ICSI 90. Piscataway, NJ, United States: Publ by IEEE. 1990. p. 752-761
Leung, P. M Perry ; Srivastava, Jaideep ; Tsai, W. T. / Performance evaluation of distributed real-time scheduling algorithms. Proc First Int Conf Syst Integr ICSI 90. Piscataway, NJ, United States : Publ by IEEE, 1990. pp. 752-761
@inproceedings{398e8701105143ce86f10e2da4145d94,
title = "Performance evaluation of distributed real-time scheduling algorithms",
abstract = "A performance evaluation of the two classes of real-time distributed CPU scheduling algorithms (sender-initiated and receiver-initiated) has been conducted, with emphasis on several factors, including network load variance, task laxity, communication costs, and mean arrival rate. Results show the pivotal role of task laxity, communication cost, and network load variance in determining the performance of each algorithm. It was observed that, when the algorithms are run on a fully connected network, receiver-initiated algorithms outperform sender-initiated algorithms with low-laxity tasks. If the laxity of the tasks is high, then both sender-initiated and receiver-initiated perform well, and the difference is not significant. Due to the low communication overhead on a fully connected network, the receiver initiated algorithm can take advantage of a frequent polling scheme to maintain up-to-date peers' CPU surplus information. Consequently, it achieves a relatively low overall miss ratio.",
author = "Leung, {P. M Perry} and Jaideep Srivastava and Tsai, {W. T.}",
year = "1990",
language = "English",
isbn = "0818690275",
pages = "752--761",
booktitle = "Proc First Int Conf Syst Integr ICSI 90",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - Performance evaluation of distributed real-time scheduling algorithms

AU - Leung, P. M Perry

AU - Srivastava, Jaideep

AU - Tsai, W. T.

PY - 1990

Y1 - 1990

N2 - A performance evaluation of the two classes of real-time distributed CPU scheduling algorithms (sender-initiated and receiver-initiated) has been conducted, with emphasis on several factors, including network load variance, task laxity, communication costs, and mean arrival rate. Results show the pivotal role of task laxity, communication cost, and network load variance in determining the performance of each algorithm. It was observed that, when the algorithms are run on a fully connected network, receiver-initiated algorithms outperform sender-initiated algorithms with low-laxity tasks. If the laxity of the tasks is high, then both sender-initiated and receiver-initiated perform well, and the difference is not significant. Due to the low communication overhead on a fully connected network, the receiver initiated algorithm can take advantage of a frequent polling scheme to maintain up-to-date peers' CPU surplus information. Consequently, it achieves a relatively low overall miss ratio.

AB - A performance evaluation of the two classes of real-time distributed CPU scheduling algorithms (sender-initiated and receiver-initiated) has been conducted, with emphasis on several factors, including network load variance, task laxity, communication costs, and mean arrival rate. Results show the pivotal role of task laxity, communication cost, and network load variance in determining the performance of each algorithm. It was observed that, when the algorithms are run on a fully connected network, receiver-initiated algorithms outperform sender-initiated algorithms with low-laxity tasks. If the laxity of the tasks is high, then both sender-initiated and receiver-initiated perform well, and the difference is not significant. Due to the low communication overhead on a fully connected network, the receiver initiated algorithm can take advantage of a frequent polling scheme to maintain up-to-date peers' CPU surplus information. Consequently, it achieves a relatively low overall miss ratio.

UR - http://www.scopus.com/inward/record.url?scp=0025567460&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025567460&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0818690275

SP - 752

EP - 761

BT - Proc First Int Conf Syst Integr ICSI 90

PB - Publ by IEEE

CY - Piscataway, NJ, United States

ER -