Scalable root cause analysis assisted by classified alarm information model based algorithm

One of the important issues for telecom carrier is that the time required to identify the root cause of a failure has increased since the number and types of alarms caused by network or service failures has increased in a fixed-mobile convergence network environment,. To address this issue, this paper proposes a root cause analysis (RCA) mechanism which classifies alarms based on their types of failures, such as resource, performance and service failures, and then promptly identifies the root cause by using a hierarchical alarm information model. Our proposed mechanism which is the implemented into a prototype system was successfully demonstrated in a testbed. Its effectiveness was validated that our RCA mechanism handled 65,000 alarms within 550 seconds in a practical network consisting of 100,000 equipments. The results also show that the algorithm minimizes the overhead of RCA itself to apply large-scale environment, and thus the total RCA performance is limited only by the DB access.