Sparse networks tolerating random faults

  • Authors:
  • T. Yamada;K. Nomura;S. Ueno

  • Affiliations:
  • Department of Information and Computer Sciences, Saitama University, Saitama 338-8570, Japan;Department of Communications and Integrated Systems, Graduate School of Science and Engineering, Tokyo Institute of Technology, Tokyo 152-8552, Japan;Department of Communications and Integrated Systems, Graduate School of Science and Engineering, Tokyo Institute of Technology, Tokyo 152-8552, Japan

  • Venue:
  • Discrete Applied Mathematics - Special issue on international workshop on algorithms, combinatorics, and optimization in interconnection networks (IWACOIN '99)
  • Year:
  • 2004

Quantified Score

Hi-index 0.01

Visualization

Abstract

A network G* is called random-fault-tolerant (RFT) network for a network G if G* contains a fault-free isomorphic copy of G with high probability even if each processor fails independently with constant probability. This paper proposes a general method to construct an RFT network G* for any network G with N processors such that. G* has O(N) processors. Based on the construction, we also show that if G is a Cayley, de-Bruijn, shuffle-exchange, or partial k-tree network with N processors and M communication links then we can construct an RFT network for G with O(N) processors and O(M log N) communication links. Cayley networks contain many popular networks such as circulant, hypercube, CCC, wrapped butterfly, star, and pancake networks.