Fast fault localization mechanism based on minimum dominating set clustering in WDM networks

Several out-of-band mechanisms have been proposed as effective approaches for link fault localization in WDM networks, such as monitoring cycle, monitoring tree and monitoring trail. However, such technologies need extra resources, including dedicated supervisor lightpaths and monitors. On the contrary, an in-band mechanism called limited-perimeter vector matching fault localization protocol (LVM) is proposed to localize single link failure with low overhead using traffic lightpaths. Nevertheless, LVM leads to the high dependence of traffic lightpaths and the long fault localization time. In order to solve these issues, a fast fault localization mechanism based on Minimum Dominating Set Clustering (MDSC) is proposed in this paper. MDSC consists of two phases: the cluster allocation algorithm using minimum dominating set (CAM) and the fault localization algorithm based on clustering (FLC). According to CAM, the network is reasonably made up of a lot of clusters, and a sink node will then be selected. According to FLC, a failure can be effectively localized by a cluster headers or the sink node by collecting the status of traffic lightpaths. Theoretical analysis and simulation results demonstrate that the proposed mechanism can efficiently reduce the dependence on traffic lightpaths and optimize the speed of fault localization.