A dynamically adaptive hybrid algorithm for scheduling lightpaths in lambda-grids

  • Authors:
  • N. R. Kaushik;S. M. Figueira

  • Affiliations:
  • Dept. of Comput. Eng., Santa Clara Univ., CA, USA;Dept. of Comput. Eng., Santa Clara Univ., CA, USA

  • Venue:
  • CCGRID '05 Proceedings of the Fifth IEEE International Symposium on Cluster Computing and the Grid - Volume 01
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper focuses on the scheduling of multiple-wavelength lightpaths and outlines an algorithm to provide increased simultaneous allocations of lightpaths on the lambda-grid. Several approaches for lightpath scheduling have been proposed, and most of them involve some variation of concentrating or balancing wavelengths over the edge-disjoint paths between the source and destination specified in a request. We have used our LRSS, lightpath request scheduling simulator, to show that the efficiency of balancing- and concentrating-based algorithms depends on the characteristics of the lightpaths requested. Based on this assumption, we have developed a dynamically adaptive hybrid algorithm, which combines both balancing and concentrating, employing each of the approaches when appropriate. The efficiency of the new algorithm is assessed by comparing its performance with the performance of the two basic approaches. Our experiments show that the dynamically adaptive hybrid scheduling algorithm has a blocking probability that is always close to the lower one, for a representative set of topologies under a representative set of traffic conditions.