Sequential Proximity-Based Clustering for Telecommunication Network Alarm Correlation
ISNN '08 Proceedings of the 5th international symposium on Neural Networks: Advances in Neural Networks, Part II
Behavioural Proximity Discovery: an adaptive approach for root cause analysis
International Journal of Business Intelligence and Data Mining
Hi-index | 0.00 |
In this paper we introduce an efficient clustering algo- rithm embedded in a novel approach for solving the prob- lem of faults identification in large telecommunication net- works. Our algorithm is especially designed for the event correlation problem taking into account comprehensive in- formation about the system behaviour. Although alarms are usually useful for identifying faults in such systems, their large number overloads the current management systems, making it extremely difficult to provide an answer within a reasonable response time. The alarm flow presents some interesting characteristics like alarm storm and alarm cas- cade. For instance, a single fault may result in a large num- ber of alarms, and it is often very difficult to isolate the true cause of the fault. One way of overcoming this problem is to analyze, interpret and reduce the number of these alarms before trying to localize the faults. In this paper, we present F ECk, and we compare it with some available clustering algorithms by experimenting them with some samples from both simulated and real data from Ericsson's network.