Photonic Network Communications

, Volume 17, Issue 1, pp 11–20

A new lightpath establishing method for dynamic traffic grooming under the overlay model

  • Bin Chen
  • Sanjay Kumar Bose
  • Wen-De Zhong
  • Hui Wang
Article

DOI: 10.1007/s11107-008-0139-9

Cite this article as:
Chen, B., Bose, S.K., Zhong, WD. et al. Photon Netw Commun (2009) 17: 11. doi:10.1007/s11107-008-0139-9

Abstract

This paper proposes the use of the saturated cut (SC) method to improve the network performance through efficient dynamic traffic grooming in IP/MPLS over WDM networks using an overlay model. With the SC information from the IP/MPLS layer, the optical layer can calculate a lightpath between two node sets rather than the earlier suggested approach of calculating such paths just between two nodes for better performance. We have used two criteria, bandwidth and path length, to find SCs and have proposed two algorithms which incorporate this in the path inflation control (PIC) policy proposed earlier. Two typical networks, NSF and COST239, were used for examining the network performance. The simulation results show that though both methods can significantly improve the overall network performance the SC method is more efficient for a sparse network like the NSF network rather than for a dense network like the COST239 network. Since calculating lightpaths for all node pairs for a SC is time consuming, we propose a considerably simpler heuristic approach that can provide almost the same level of performance improvement as the SC method but with much lower complexity.

Keywords

IP/MPLS over WDM Traffic grooming Saturated cut method Light path LSP 

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • Bin Chen
    • 1
  • Sanjay Kumar Bose
    • 2
  • Wen-De Zhong
    • 2
  • Hui Wang
    • 1
  1. 1.Faculty of Information EngineeringShenzhen UniversityShenzhenPeople’s Republic of China
  2. 2.School of Electrical and Electronic Engineering, Network Technology Research CentreNanyang Technological UniversitySingaporeSingapore

Personalised recommendations