Premium
Some analysis of call blocking probabilities in hierarchical WDM networks offering multiclass service
Author(s) -
Kuppuswamy Kalyan,
Lee Daniel C.
Publication year - 2011
Publication title -
international journal of communication systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.344
H-Index - 49
eISSN - 1099-1131
pISSN - 1074-5351
DOI - 10.1002/dac.1177
Subject(s) - computer science , blocking (statistics) , traverse , call blocking , computer network , hierarchy , node (physics) , wavelength division multiplexing , distributed computing , routing (electronic design automation) , service (business) , telecommunications network , hierarchical routing , hierarchical network model , network topology , quality of service , routing protocol , static routing , wavelength , physics , economy , optoelectronics , geodesy , structural engineering , economics , engineering , market economy , geography
We envision a large‐scale multiclass optical network in which lightpaths are dynamically set up and terminated between source and destination nodes and these lightpaths traverse through a number of interconnected WDM networks in different administrative domains. In this envisioned multiclass network, calls are classified by the number of wavelengths demanded per call and the statistics of call holding times (or service subscription periods). We present an approximation method for the performance analysis of such a large‐scale multiclass optical WDM network. We model the large‐scale optical network as a two‐level hierarchical multiclass loss network; the lower level of hierarchy consists of individual optical WDM networks (network segments). Each of these network segments is abstracted to a logical node at the higher level of hierarchy, and these logical nodes are connected to each other through logical links. We also assume that the call processing mechanism in this large‐scale optical network resorts to hierarchical routing. The algorithms presented in this paper compute per‐class end‐to‐end approximate blocking probabilities . Our experiments show that the simulation results match closely with the results of the proposed analytical approximation methods, thus validating the proposed methodology. Copyright © 2010 John Wiley & Sons, Ltd.