Optimization models for ring-planning decision support

摘要

Today, telecommunications providers are experiencing an unprecedented growth in demand for bandwidth capacity. Especially as we look to the future,this demand will be satisfied using new optical networking technology, featuring Dense Wavelength Division Multiplexing (DWDM) transmission systems over fiber and Intelligent Optical Switches (IOSs). These systems promise capabilities for rapid, remote provisioning of services. In particular, suing IOSs, the provisioning of ring structures can be simplified. Such ring structures can provide rapid restoration capabilities in case of a fiber cut. In this paper, we describe optimization models that can be used for capacity planning in ring-based broad-hand optical networks. Given and anticipated set of demands and planned routings, the basic model can be used to choose a best set of rings and to determine the number of wavelengths required on each fiber link. Variations on the basic model can consider cases in which the number of ring topologies chosen must be limited, or in which there are limited wavelengths that can be made available on various links, etc. We discuss how these models can be sued in various types of case studies, e.g., to evaluate different possible sets of ring families, to determine robust capacity augments, or to evaluate different ring reconfiguration strategies. The basic model and its variations are primarily intended for used in long-range planning. However, we show how the solution methodology for the basic model can also be incorporated within a larger, heuristic framework, in order to address more short-term provisioning issues. This larger framework must give particular consideration to the amount of available capacity and how demand is routed on available wavelengths.