Lightpath topology configuration for wavelength-routed IP/MPLS networks in a multi-layered environment: Integrated models, algorithms, and analysis.

摘要

In this dissertation, we explore the problem of optimal lightpath topology configuration and efficient traffic engineering of IP traffic in wavelength-routed optical networks using a multi-layered environment.;First, we consider the shortest path routing framework such as OSPF/IS-IS for IP networks; in particular, we consider non-bifurcated minimum-hop routing to route IP traffic on the lightpath topology. We present a novel model and solve the problem of lightpath topology design and traffic engineering such that all demand pairs have unique-shortest path in the lightpath topology; our approach specifically addresses design in the presence of limited number of transreceivers.;To incorporate the survivability requirement such that the lightpath topology remains connected in the event of any single physical link failure, we have extended the above model. In particular, we present a novel solution approach, which designs a unique shortest path-based lightpath topology with the maximum number of links that are part of at least one cycle and provide survivable mapping of the lightpath topology on the physical topology. We use "total demand volume that can be protected in the lightpath topology" as a measure to evaluate the effectiveness of our solution approach in generating minimum average packet hop-based survivable lightpath topologies.;Next, we consider the lightpath topology configuration problem for time-varying traffic in an MPLS routing environment. We propose two approaches to solve the problem: static and dynamic configuration. For static configuration, we present a model to design a single lightpath topology configuration optimized for multiple traffic matrices. The model considers the joint problem of optimal configuration of lightpath topology and optimal routing of LSPs.;Finally, for the dynamic configuration, we present a solution approach based on Lagrangian Relaxation and sub-gradient optimization, where the major thrust is to design a lightpath topology optimized for new traffic matrix with fewest possible number of lightpath deletions from the current lightpath topology. Through our studies, we have examined the trade-off between the number of allowed lightpath disruptions and the average packet hop-distance value.