Failure location in WDM networks

Abstract

Fault identification and location in optical networks must cope with a multitude of factors: (i) the redundancy and the lack of coordination (internetworking) of the managements at the different layers (WDM, SDH/SONET, ATM, IP); (ii) the large number of alarms a single failure can trigger; (iii) the difficulty in detecting some failures and the resulting need to cope with missing or false alarms. This chapter first details the behavior of network components in transparent WDM networks when a failure occurs. Using this model, we then describe an efficient algorithm (Fault Location Algorithm, FLA) pointing out the element(s) which is (are) most likely to be the cause of the received alarms. Although the problem of multiple failure diagnosis is known to be NP-hard, the non-polynomial complexity of the algorithm is pushed ahead in a pre-computational phase, which can be done off-line, and not at the time of a failure. The diagnosis phase is therefore very rapid. We discuss the time and space complexity of the FLA.