Resource Assignment for Fiber Optic Networks

Transport optical networks form the backbone of the information infrastructure worldwide. Current systems use wavelength division multiplexing (WDM) technology to accommodate large traffic volumes. In the near future elastic optical networks (EON) are expected to replace WDM networks to further increase the network capacity. This dissertation examines the problem of assigning physical and spectral resources to WDM networks and EON for efficient design and use of these systems.

The physical resource assignment problem, often referred to as the routing and wavelength assignment (RWA) problem, is very important part of the design of fiber optic WDM networks. Objectives of this problem include minimizing the total capital investment in the static design phase and maximizing the throughput in the dynamic operation phase. We develop strategies both in heuristic algorithms and in mixed-integer linear programming (MILP) not just for RWA but also to include considerations of physical impairments, traffic grooming for both static and dynamic networks. For heuristic algorithms, we develop both centralized and distributed algorithms based on the information sharing and assignment decision making. The distributed heuristic algorithms are based on ant colony optimization (ACO) which is a meta-heuristic method that is inspired by the foraging behavior of ants and has been widely implemented in solving discrete optimization problems. Simulation results show although the centralized algorithm shows better efficiency in terms of blocking probability, our ACO shows great robustness and adaptivity to varying network and traffic conditions. We also show implementing technologies such as traffic grooming and signal regeneration will greatly reduce the blocking probability of calls.

Elastic optical networks (EON) have added flexibility to network deployment and management. We propose a link-based MILP formulation for EON to implement signal regeneration as well as wavelength conversion and modulation conversion. We then propose a recursive model in order to either augment existing network deployments or speed up the resource allocation computation time for larger networks with higher traffic demand requirements than can be solved using an MILP. We show through simulation that systems equipped with signal regenerators or wavelength converters require a notably smaller total bandwidth, depending on the topology of the network. We also show that the suboptimal recursive solution speeds up the calculation and makes the running-time more predictable, compared to the optimal MILP. We compare the two approaches, namely path-based (PB) and link-based (LB) MILP formulations, in their implementation, optimality, and complexity for EONs. We show using simulation that it is beneficial to use LB formulation when including the signal regeneration and that the network topology and traffic demand affect the difference in performances between the two formulations.
We combine the MILP formulation for static network and time-slot concept to solve a real-time traffic scenario so the overall network throughput is maximized. The throughput is maximized given the current network state.

Impact of technologies such as signal regeneration, wavelength conversion, and modulation conversion on network performance metrics such as the amount of spectrum needed is also investigated by analytical modeling. Analytical modeling provides the desirable method for network designers to begin with a fast coarse estimate of network performance implementing such technologies without requiring computationally-burdensome and detailed algorithms.