Integer and combinatorial optimization
Integer and combinatorial optimization
Routing and wavelength assignment in all-optical networks
IEEE/ACM Transactions on Networking (TON)
Balancing loads on SONET rings with integer demand splitting
Computers and Operations Research
SONET ring sizing with genetic algorithms
Computers and Operations Research
SIAM Journal on Discrete Mathematics
Fiber Network Service Survivability
Fiber Network Service Survivability
Multiwavelength Optical Networks with Limited Wavelength Conversion
INFOCOM '97 Proceedings of the INFOCOM '97. Sixteenth Annual Joint Conference of the IEEE Computer and Communications Societies. Driving the Information Revolution
WDM: North American deployment trends
IEEE Communications Magazine
Current European WDM deployment trends
IEEE Communications Magazine
Wave-division multiplexing in the Sprint long distance network
IEEE Communications Magazine
IEEE Communications Magazine
Management and control of transparent optical networks
IEEE Journal on Selected Areas in Communications
| Hi-index | 0.00 |
We address two questions related to the design and management of ring-based networks and the choice of add–drop multiplexer (ADM) technology. One issue is the benefit of a channel interchange capability in the ADMs of SONET or WDM rings. Although counterintuitive, we show that there is essentially no benefit from this potentially expensive capability if the selection of demands for ring loading is optimally planned for each case. This knowledge can help prioritize standards-making and equipment development efforts. A second question relates to provisioning demands over more than one route through a ring. Administration and provisioning is simplified by prescribing a single route per origin–destination pair, but in a ring-based network there is an as yet unquantified penalty to enforcing this operational policy rather than allowing “splitting” of routes within rings. To assess the benefits of channel interchange and route splitting, together and individually, we solve the respective integer programming formulation to find the optimal revenue or transport efficiency loading solutions. Conclusions are based on comparing efficiency and revenue metrics on a large number of randomized trials with different ring sizes, demand patterns, and line capacities. Results show that while channel interchange has negligible beneficial effect, route splitting can provide significant revenue or efficiency benefits.