Overview

Any network is subject to failures, whether it be due to fiber cuts, equipment failures, software errors, technician errors, or environmental causes. Protection against failures, by providing alternative paths or backup equipment, is a necessary component of network design. One of the key design decisions is selecting the networking layer, or layers, in which to implement protection. For example, higher-layer protocols, such as IP, typically have well defined protection mechanisms. However, these mechanisms usually operate on a relatively fine traffic granularity; as traffic levels increase, implementing failure recovery solely in these layers may be too slow. Optical protection, which operates on the granularity of a wavelength (or even a waveband or a fiber), has received growing attention, largely due to its ability to scale more gracefully with increasing traffic levels. As the wavelength bit-rate increases, e.g., from 10 Gb/s to 40 Gb/s and beyond, the amount of network traffic that can be restored by rerouting a wavelength grows accordingly.

There are numerous optical protection schemes, where the mechanisms differ in the amount of spare capacity and equipment required, the speed of protection, the number of simultaneous failures from which recovery is possible, and the operational complexity. It is possible to support a combination of protection mechanisms in a network, where the protection scheme used for a particular demand depends in large part on the associated availability requirements. Such requirements are usually specified as part of a service level agreement (SLA) between a carrier and its customers. For example, some...