Support Efficient and Fault-Tolerant Multicast in Bufferless Network-on-Chip

Chaochao FENG; Zhonghai LU; Axel JANTSCH; Minxuan ZHANG; Xianju YANG

doi:10.1587/transinf.E95.D.1052

Support Efficient and Fault-Tolerant Multicast in Bufferless Network-on-Chip

Chaochao FENG, Zhonghai LU, Axel JANTSCH, Minxuan ZHANG, Xianju YANG

Full Text Views

0

Share
Cite this

Summary :

In this paper, we propose three Deflection-Routing-based Multicast (DRM) schemes for a bufferless NoC. The DRM scheme without packets replication (DRM_noPR) sends multicast packet through a non-deterministic path. The DRM schemes with adaptive packets replication (DRM_PR_src and DRM_PR_all) replicate multicast packets at the source or intermediate node according to the destination position and the state of output ports to reduce the average multicast latency. We also provide fault-tolerant supporting in these schemes through a reinforcement-learning-based method to reconfigure the routing table to tolerate permanent faulty links in the network. Simulation results illustrate that the DRM_PR_all scheme achieves 41%, 43% and 37% less latency on average than that of the DRM_noPR scheme and 27%, 29% and 25% less latency on average than that of the DRM_PR_src scheme under three synthetic traffic patterns respectively. In addition, all three fault-tolerant DRM schemes achieve acceptable performance degradation at various link fault rates without any packet lost.

Publication: IEICE TRANSACTIONS on Information Vol.E95-D No.4 pp.1052-1061

Publication Date: 2012/04/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E95.D.1052

Type of Manuscript: PAPER

Category: Computer System

Cite this

Copy

Chaochao FENG, Zhonghai LU, Axel JANTSCH, Minxuan ZHANG, Xianju YANG, "Support Efficient and Fault-Tolerant Multicast in Bufferless Network-on-Chip" in IEICE TRANSACTIONS on Information, vol. E95-D, no. 4, pp. 1052-1061, April 2012, doi: 10.1587/transinf.E95.D.1052.
Abstract: In this paper, we propose three Deflection-Routing-based Multicast (DRM) schemes for a bufferless NoC. The DRM scheme without packets replication (DRM_noPR) sends multicast packet through a non-deterministic path. The DRM schemes with adaptive packets replication (DRM_PR_src and DRM_PR_all) replicate multicast packets at the source or intermediate node according to the destination position and the state of output ports to reduce the average multicast latency. We also provide fault-tolerant supporting in these schemes through a reinforcement-learning-based method to reconfigure the routing table to tolerate permanent faulty links in the network. Simulation results illustrate that the DRM_PR_all scheme achieves 41%, 43% and 37% less latency on average than that of the DRM_noPR scheme and 27%, 29% and 25% less latency on average than that of the DRM_PR_src scheme under three synthetic traffic patterns respectively. In addition, all three fault-tolerant DRM schemes achieve acceptable performance degradation at various link fault rates without any packet lost.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.E95.D.1052/_p

Copy

@ARTICLE{e95-d_4_1052,
author={Chaochao FENG, Zhonghai LU, Axel JANTSCH, Minxuan ZHANG, Xianju YANG, },
journal={IEICE TRANSACTIONS on Information},
title={Support Efficient and Fault-Tolerant Multicast in Bufferless Network-on-Chip},
year={2012},
volume={E95-D},
number={4},
pages={1052-1061},
abstract={In this paper, we propose three Deflection-Routing-based Multicast (DRM) schemes for a bufferless NoC. The DRM scheme without packets replication (DRM_noPR) sends multicast packet through a non-deterministic path. The DRM schemes with adaptive packets replication (DRM_PR_src and DRM_PR_all) replicate multicast packets at the source or intermediate node according to the destination position and the state of output ports to reduce the average multicast latency. We also provide fault-tolerant supporting in these schemes through a reinforcement-learning-based method to reconfigure the routing table to tolerate permanent faulty links in the network. Simulation results illustrate that the DRM_PR_all scheme achieves 41%, 43% and 37% less latency on average than that of the DRM_noPR scheme and 27%, 29% and 25% less latency on average than that of the DRM_PR_src scheme under three synthetic traffic patterns respectively. In addition, all three fault-tolerant DRM schemes achieve acceptable performance degradation at various link fault rates without any packet lost.},
keywords={},
doi={10.1587/transinf.E95.D.1052},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - Support Efficient and Fault-Tolerant Multicast in Bufferless Network-on-Chip
T2 - IEICE TRANSACTIONS on Information
SP - 1052
EP - 1061
AU - Chaochao FENG
AU - Zhonghai LU
AU - Axel JANTSCH
AU - Minxuan ZHANG
AU - Xianju YANG
PY - 2012
DO - 10.1587/transinf.E95.D.1052
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E95-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2012
AB - In this paper, we propose three Deflection-Routing-based Multicast (DRM) schemes for a bufferless NoC. The DRM scheme without packets replication (DRM_noPR) sends multicast packet through a non-deterministic path. The DRM schemes with adaptive packets replication (DRM_PR_src and DRM_PR_all) replicate multicast packets at the source or intermediate node according to the destination position and the state of output ports to reduce the average multicast latency. We also provide fault-tolerant supporting in these schemes through a reinforcement-learning-based method to reconfigure the routing table to tolerate permanent faulty links in the network. Simulation results illustrate that the DRM_PR_all scheme achieves 41%, 43% and 37% less latency on average than that of the DRM_noPR scheme and 27%, 29% and 25% less latency on average than that of the DRM_PR_src scheme under three synthetic traffic patterns respectively. In addition, all three fault-tolerant DRM schemes achieve acceptable performance degradation at various link fault rates without any packet lost.
ER -