Networking

Efficient and Robust Congestion Estimation for Dynamic WDM Networks

Free registration required

Executive Summary

Emerging applications require high-bandwidth network connectivity on demand, exacerbating the already bursty nature of data traffic and motivating the development of new dynamic management techniques for the wavelength based optical backbones underlying the Internet. This paper focuses on the problem of routing new connection requests in such an environment, introducing efficient algorithms as well as models that help to provide a deeper understanding of the algorithms' benefits. The authors first introduce a threshold-based Congestion Aware Routing (CAR) algorithm that leverages quick congestion estimation and admission control techniques to dynamically shift between low- and high-load strategies. Using the link load information from a prospective path, the congestion estimator responds effectively to localize congestions.

  • Format: PDF
  • Size: 415.9 KB