M2-CYCLE: An optical layer algorithm for fast link failure detection in all-optical mesh networks

  • Authors:
  • Bin Wu;Kwan L. Yeung;Bing Hu;Pin-Han Ho

  • Affiliations:
  • Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, ON, Canada N2L 3G1;Department of Electrical and Electronic Engineering, The University of Hong Kong, Pokfulam, Hong Kong;Department of Information Science and Electronic Engineering, Zhejiang University, Hangzhou, China;Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, ON, Canada N2L 3G1

  • Venue:
  • Computer Networks: The International Journal of Computer and Telecommunications Networking
  • Year:
  • 2011

Quantified Score

Hi-index 0.01

Visualization

Abstract

To achieve fast link failure detection in all-optical networks, monitoring-cycles (m-cycles) are introduced at the optical layer to reduce the number of required monitoring devices (or monitors). Each m-cycle is equipped with a monitor and a pair of optical transceivers to transmit an optical supervisory signal. A set of m-cycles can be found to form a cycle cover of the network. If a link fails, optical supervisory signals inside the m-cycles passing through this link will be disrupted, and the corresponding monitors will alarm due to Loss of Light (LoL). This gives an alarm code to localize the failed link. The accuracy of the failure localization is measured by localization degree, and the amount of monitoring resources required is measured by the number of cycles/monitors, cover length, and monitoring wavelength requirement. The best known m-cycle construction algorithm HST [11] adopts a spanning tree-based approach. In this paper, we propose a new algorithm M^2-CYCLE to construct a cycle cover consisting of a set of minimum-length m-cycles (or m^2-cycles). We prove that M^2-CYCLE achieves the same localization degree as the spanning tree-based approach, but requires less amount of monitoring resources no matter how the spanning tree is generated. Numerical results confirm our theoretical analysis, and show that the monitoring resources required by M^2-CYCLE are dramatically cut down.