Abstract

Edge routers are one of the key elements in an optical burst switched (OBS) network. Packets are assembled into bursts and then disassembled after being switched through the network. In a span-constrained OBS network,the edge router latency dominates propagation delay through the network. Consequently,the analysis of the edge router latency under different input traffic conditions and for different edge router parameters is essential for achieving an optimized network performance. In this paper, we present a framework to model and analyze in detail the performance of an edge router for span constrained networks under varying traffic utilizations, traffic statistics and burst assembly parameters. Two burst assembly algorithms are presented, and their impact on burst size and assembly time is studied for fixed and variable-size packets. We identify three distinct regions of edge router operation as a function of the input traffic statistics and burst assembly parameters. This enabled us to better understand the behavior of edge router latency under different input traffic conditions. The results of the analysis are then used to formulate design principles for the edge router. The impact of traffic self-similarity in each of these regions of operation was also analyzed. It is shown that self-similar traffic affects the variance of the burst size and assembly time, but does not impact overall edge router latency.

© 2004 IEEE

PDF Article

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription