| To handle the increasing amount and heterogeneity of Internet trac, a set of recent innovations have been proposed for optical backbone networks. One is mixed-line-rate (MLR) network, where the wavelength channels of optical paths (i.e., light-paths) can have a variety of capacities (10/40/100 Gbps) in a wavelength-division multiplexing (WDM) network. Another is elastic network, where the traditional ITU-T wavelength grid is not followed, but appropriate-sized optical bandwidth in sub- (e.g., 100 Gbps) granularity is allocated to an end-to-end path. In both networks, survivability is very important, because the failure of a network element (e.g., a fiber cut) can cause the failure of several lightpaths, resulting in huge data loss, especially if the lightpaths are migrated to high bit rates, such as 100 Gbps and beyond, which are expected to be accommodated in MLR and future elastic networks. Hence, we study the protection strategies in these two new optical network paradigms.;We propose to design transparent MLR networks with dedicated and shared protection. Under dedicated protection, a lightpath is protected by a link-disjoint lightpath, with the same or different rates. Under shared protection, a lightpath is protected by a series of pre-lit lightpaths. When a failure occurs on a working lightpath, these pre-deployed lightpaths are concatenated to form a backup path. Working lightpaths can share capacity on common backup lightpaths if they are link disjoint.;We also propose a novel shared-protection approach specific to elastic networks. It not only provides traditional backup sharing, but it also offers a new opportunity of spectrum sharing enabled by the elasticity of the transponders: 1) if the working paths of two connections are link disjoint physically, and 2) if their backup paths traverse two lightpaths which are adjacent on a fiber link, then the two backup lightpaths can share spectrum.;Therefore, this dissertation proposes new and improved strategies for protection in new optical networking paradigms. To the best of our knowledge, it is among the first studies to explore the opportunities that MLR and elastic networks enable the support of their effective protection. |
