Unidirectional star graphs

Khaled Day, Anand Tripathi

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

We propose an assignment of directions to the edges of the n-star graph and derive attractive properties for the resulting unidirectional star graph USn. A simple polarity function is used to define the directions of the edges. USn is shown to be strongly connected and recursively structured. The number of incoming ports of a node in USn is equal to the number of its outgoing ports (plus or minus one when n is even). The polarities of the ports of a neighboring node can be directly obtained from the polarities of the ports of the local node. These properties have allowed the design of a near optimal distributed routing algorithm as indicated by simulation results on all USn graphs with up to 362,880 nodes.

Original languageEnglish
Pages (from-to)123-129
Number of pages7
JournalInformation Processing Letters
Volume45
Issue number3
DOIs
Publication statusPublished - Mar 8 1993

Fingerprint

Star Graph
Stars
Polarity
Routing algorithms
Vertex of a graph
Parallel algorithms
Routing Algorithm
Distributed Algorithms
Assignment
Graph in graph theory
Simulation

Keywords

  • Combinatorial problems
  • directed graphs
  • distributed computing
  • distributed routing
  • interconnection networks
  • star graphs

ASJC Scopus subject areas

  • Computational Theory and Mathematics

Cite this

Unidirectional star graphs. / Day, Khaled; Tripathi, Anand.

In: Information Processing Letters, Vol. 45, No. 3, 08.03.1993, p. 123-129.

Research output: Contribution to journalArticle

Day, Khaled ; Tripathi, Anand. / Unidirectional star graphs. In: Information Processing Letters. 1993 ; Vol. 45, No. 3. pp. 123-129.
@article{ea6302487c8d40d7b9c42fd616097be3,
title = "Unidirectional star graphs",
abstract = "We propose an assignment of directions to the edges of the n-star graph and derive attractive properties for the resulting unidirectional star graph USn. A simple polarity function is used to define the directions of the edges. USn is shown to be strongly connected and recursively structured. The number of incoming ports of a node in USn is equal to the number of its outgoing ports (plus or minus one when n is even). The polarities of the ports of a neighboring node can be directly obtained from the polarities of the ports of the local node. These properties have allowed the design of a near optimal distributed routing algorithm as indicated by simulation results on all USn graphs with up to 362,880 nodes.",
keywords = "Combinatorial problems, directed graphs, distributed computing, distributed routing, interconnection networks, star graphs",
author = "Khaled Day and Anand Tripathi",
year = "1993",
month = "3",
day = "8",
doi = "10.1016/0020-0190(93)90013-Y",
language = "English",
volume = "45",
pages = "123--129",
journal = "Information Processing Letters",
issn = "0020-0190",
publisher = "Elsevier",
number = "3",

}

TY - JOUR

T1 - Unidirectional star graphs

AU - Day, Khaled

AU - Tripathi, Anand

PY - 1993/3/8

Y1 - 1993/3/8

N2 - We propose an assignment of directions to the edges of the n-star graph and derive attractive properties for the resulting unidirectional star graph USn. A simple polarity function is used to define the directions of the edges. USn is shown to be strongly connected and recursively structured. The number of incoming ports of a node in USn is equal to the number of its outgoing ports (plus or minus one when n is even). The polarities of the ports of a neighboring node can be directly obtained from the polarities of the ports of the local node. These properties have allowed the design of a near optimal distributed routing algorithm as indicated by simulation results on all USn graphs with up to 362,880 nodes.

AB - We propose an assignment of directions to the edges of the n-star graph and derive attractive properties for the resulting unidirectional star graph USn. A simple polarity function is used to define the directions of the edges. USn is shown to be strongly connected and recursively structured. The number of incoming ports of a node in USn is equal to the number of its outgoing ports (plus or minus one when n is even). The polarities of the ports of a neighboring node can be directly obtained from the polarities of the ports of the local node. These properties have allowed the design of a near optimal distributed routing algorithm as indicated by simulation results on all USn graphs with up to 362,880 nodes.

KW - Combinatorial problems

KW - directed graphs

KW - distributed computing

KW - distributed routing

KW - interconnection networks

KW - star graphs

UR - http://www.scopus.com/inward/record.url?scp=0027911739&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027911739&partnerID=8YFLogxK

U2 - 10.1016/0020-0190(93)90013-Y

DO - 10.1016/0020-0190(93)90013-Y

M3 - Article

AN - SCOPUS:0027911739

VL - 45

SP - 123

EP - 129

JO - Information Processing Letters

JF - Information Processing Letters

SN - 0020-0190

IS - 3

ER -