Monitoring cycles for fault detection in meshed all-optical networks

Fault detection is critical for all-optical networks (AONs). This paper introduces the concept of monitoring cycle and proposes a fault detection mechanism based on decomposing AONs into a set of cycles (a cycle cover), in which each one is defined as a monitoring cycle. Two cycle-finding algorithms are developed and compared for the proposed fault detection mechanism: heuristic depth first searching (HDFS) and shortest path Eulerian matching (SPEM). The degradation of wavelength utilization and the cardinality of cycle covers are analyzed for evaluating the proposed mechanism. The proposed mechanism is applied to four network examples: NSFNET, ARPA2, SmallNet and Bellcore. The evaluation results show that the proposed fault detection mechanism is effective and cost efficient.