Reliability of the Internet became a critical issue in the past years. The permanent monitoring, surveillance of the optical backbone networks along with failure detection, localization and correction functions are fundamental requirements, that an operator must satisfy. The traditional approach for failure detection and alarm message dissemination uses control plane packets, which generates a significant delay. A possible solution for implementing optical layer failure localization in the network is using multihop supervisory lightpaths, called monitoring trails. These trails intrinsically carry a single bit of information, representing the statuses of the network elements they traverse by. By properly defining and utilizing monitoring trails, all nodes can unambiguously identify the faulty element in the network, almost instantaneously.
In this study a number of constructions with optimal, or near-optimal solutions for the node failure localization problem on certain special topologies are presented. Moreover, a heuristic algorithm is being introduced, that achieves network-wide unambiguous failure localization in optical networks. The proposed recursive algorithm had been validated on a large number of topologies. Results show, that the proposed algorithm significantly outperforms the previously known best heuristic for the given problem.