A Full Automatic Identification of Faulty Links for Communication Networks

Abstract

Conventionally the manual service is often required for fault diagnosis of network links with improper accuracy brought in. So, a full automatic fault diagnosis approach for communication links is proposed as follows. If the path information that which node are available or unavailable to communicate with from the node(s) taking charge of network management has both been known, a fault diagnosis model is developed in combination with probability computation, by which the node(s) in charge of network management can implement full automatically the real-time identification of the range where the link failures happen most probably. Then, the link failures can be positioned accurately via quick tests. Based on this approach, a hierarchy network management architecture is designed to deal with the fault diagnosis for a heterogeneous network environment. The simulation results showed that this approach is real-time, highly accurate and full automatic with no or few extra bandwidth occupied.