| Optical wavelength division multiplexing (WDM) technology is being widely used in communication networks to meet the rapid growth of traffic demands. While WDM can dramatically increase the transmission capacity, up to several Terabits per second, it also leads to many challenging problems in the design and operation of optical WDM networks. In this dissertation, we consider three design and operation problems for WDM networks.; The first problem is the problem of traffic grooming for WDM ring networks. In particular, we consider the problem of traffic grooming with dynamic traffic. For a ring network, add/drop multiplexers (ADMs) make up the majority of network cost. Thus, our objective is to minimize the number of ADMs. This problem can be formulated as an integer linear programming (ILP) problem; however, solving the ILP problem is computationally infeasible except for very small rings. We propose a method based on both heuristics and ILP formulation.; The second problem is the problem of diverse routing. In optical networks, each circuit carries a large amount of traffic, and, as a result, needs to be protected to withstand various network failures. Since the diverse routing problem for optical networks is NP-complete, we propose a two-step approach based on Lagrangian relaxation, heuristics, and Floyd-Warshall algorithm. In addition, we analyze the resource requirements of two diverse routing schemes, namely, dedicated and shared path protection schemes.; The third and final problem is the problem of traffic routing and grooming for WDM mesh networks, with the objective of maximizing the overall network revenue subject to network capacity (or total capital expenditure) constraints. This problem is very similar to the 0-1 multi-commodity flow problem. We propose various heuristic methods for solving this problem.; For all three problems, numerical results are provided to demonstrate the effectiveness of our proposed methods. |
