Backup Path Classification Based on Failure Risks for Efficient Backup Path Computation

We propose a new approach exploiting the failure risk (node, link or Shared Risk Link Group) structures to enhance the backup path computation. Upon failure, our approach classifies the backup paths into two categories: operative backup paths and inoperative backup paths. An operative backup path is an active backup path which really receives traffic of some affected communications while an inoperative backup path does not receive any traffic. With the observation that only the operative backup paths really participate to the recovery procedure, we enhance the backup path computation (1) by limiting the concurrence for the protection bandwidth allocations to the operative backup paths (instead of all the active backup paths like in the classical approaches) and (2) by reducing the set of failure risks that a backup path must bypass. Simulations shows that our approach improves the protection bandwidth allocations and decreases the ratio of rejected backup paths.