Abstract
The novel core network architecture presented in this paper realizes distributed all-optical switching of
payload by partitioning the network into a number of geographically limited domains, where two-way reservations are
effective. Thus, inside each domain, loss is eliminated, while traffic from many nodes can be aggregated into single
bursts, improving efficiency. Clustered nodes contribute contiguous optical slots, which are marshaled into
composite optical frames destined for other clusters, under the guidance of a reservation-based control protocol.
The lossless aggregation of traffic from several core nodes allows the use of cost-effective bufferless all-optical
transport among the domains with electrical buffers employed at the periphery of the system. The end result is a
triple improvement in loss probabilities, efficiency, and cost. This is achieved by exploiting three features of the
architecture: the distributed switching functionality (as in early LANs when centralized switching was expensive),
localized reservations (avoiding the intolerable delays of end-to-end reservations), and a reduced number of
source-destination pairs (by means of node clustering into reservation domains).
© 2007 IEEE
PDF Article
More Like This
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 Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription