Analysis and optimization of prediction-based flow control in networks-on-chip

Umit Ogras, Radu Marculescu

Research output: Chapter in Book/Report/Conference proceedingChapter

1 Citation (Scopus)

Abstract

While networks-on-Chip (NoC) architectures may offer higher bandwidth compared to traditional bus-based communication, their performance can degrade significantly in the absence of effective flow control algorithms. This chapter presents a predictive closed-loop flow control mechanism, which is used to predict the congestion level in the network. Based on this information, the proposed scheme controls the packet injection rate at traffic sources in order to regulate the total number of packets in the network. Finally, simulations and experimental study using our FPGA prototype show that the proposed controller delivers a better performance compared to the traditional switch-to-switch flow control algorithms under various real and synthetic traffic patterns.

Original languageEnglish (US)
Title of host publicationLecture Notes in Electrical Engineering
Pages105-133
Number of pages29
Volume184
DOIs
StatePublished - 2013
Externally publishedYes

Publication series

NameLecture Notes in Electrical Engineering
Volume184
ISSN (Print)18761100
ISSN (Electronic)18761119

Fingerprint

Flow control
Switches
Field programmable gate arrays (FPGA)
Bandwidth
Controllers
Communication
Network-on-chip

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering

Cite this

Ogras, U., & Marculescu, R. (2013). Analysis and optimization of prediction-based flow control in networks-on-chip. In Lecture Notes in Electrical Engineering (Vol. 184, pp. 105-133). (Lecture Notes in Electrical Engineering; Vol. 184). https://doi.org/10.1007/978-94-007-3958-1-7

Analysis and optimization of prediction-based flow control in networks-on-chip. / Ogras, Umit; Marculescu, Radu.

Lecture Notes in Electrical Engineering. Vol. 184 2013. p. 105-133 (Lecture Notes in Electrical Engineering; Vol. 184).

Research output: Chapter in Book/Report/Conference proceedingChapter

Ogras, U & Marculescu, R 2013, Analysis and optimization of prediction-based flow control in networks-on-chip. in Lecture Notes in Electrical Engineering. vol. 184, Lecture Notes in Electrical Engineering, vol. 184, pp. 105-133. https://doi.org/10.1007/978-94-007-3958-1-7
Ogras U, Marculescu R. Analysis and optimization of prediction-based flow control in networks-on-chip. In Lecture Notes in Electrical Engineering. Vol. 184. 2013. p. 105-133. (Lecture Notes in Electrical Engineering). https://doi.org/10.1007/978-94-007-3958-1-7
Ogras, Umit ; Marculescu, Radu. / Analysis and optimization of prediction-based flow control in networks-on-chip. Lecture Notes in Electrical Engineering. Vol. 184 2013. pp. 105-133 (Lecture Notes in Electrical Engineering).
@inbook{ab566e0fe795421a86024686b566d554,
title = "Analysis and optimization of prediction-based flow control in networks-on-chip",
abstract = "While networks-on-Chip (NoC) architectures may offer higher bandwidth compared to traditional bus-based communication, their performance can degrade significantly in the absence of effective flow control algorithms. This chapter presents a predictive closed-loop flow control mechanism, which is used to predict the congestion level in the network. Based on this information, the proposed scheme controls the packet injection rate at traffic sources in order to regulate the total number of packets in the network. Finally, simulations and experimental study using our FPGA prototype show that the proposed controller delivers a better performance compared to the traditional switch-to-switch flow control algorithms under various real and synthetic traffic patterns.",
author = "Umit Ogras and Radu Marculescu",
year = "2013",
doi = "10.1007/978-94-007-3958-1-7",
language = "English (US)",
isbn = "9789400739574",
volume = "184",
series = "Lecture Notes in Electrical Engineering",
pages = "105--133",
booktitle = "Lecture Notes in Electrical Engineering",

}

TY - CHAP

T1 - Analysis and optimization of prediction-based flow control in networks-on-chip

AU - Ogras, Umit

AU - Marculescu, Radu

PY - 2013

Y1 - 2013

N2 - While networks-on-Chip (NoC) architectures may offer higher bandwidth compared to traditional bus-based communication, their performance can degrade significantly in the absence of effective flow control algorithms. This chapter presents a predictive closed-loop flow control mechanism, which is used to predict the congestion level in the network. Based on this information, the proposed scheme controls the packet injection rate at traffic sources in order to regulate the total number of packets in the network. Finally, simulations and experimental study using our FPGA prototype show that the proposed controller delivers a better performance compared to the traditional switch-to-switch flow control algorithms under various real and synthetic traffic patterns.

AB - While networks-on-Chip (NoC) architectures may offer higher bandwidth compared to traditional bus-based communication, their performance can degrade significantly in the absence of effective flow control algorithms. This chapter presents a predictive closed-loop flow control mechanism, which is used to predict the congestion level in the network. Based on this information, the proposed scheme controls the packet injection rate at traffic sources in order to regulate the total number of packets in the network. Finally, simulations and experimental study using our FPGA prototype show that the proposed controller delivers a better performance compared to the traditional switch-to-switch flow control algorithms under various real and synthetic traffic patterns.

UR - http://www.scopus.com/inward/record.url?scp=84875902022&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84875902022&partnerID=8YFLogxK

U2 - 10.1007/978-94-007-3958-1-7

DO - 10.1007/978-94-007-3958-1-7

M3 - Chapter

SN - 9789400739574

VL - 184

T3 - Lecture Notes in Electrical Engineering

SP - 105

EP - 133

BT - Lecture Notes in Electrical Engineering

ER -