The last decade has witnessed an increasing usage of data mining techniques such as clustering into many applications including network faults classification. In communication networks often large volume of network faults are generated. Even a single fault may result in large number of alarms causing information redundancy. Providing a coherent classification scheme would certainly help network management process, avoid system breakdown, by isolating faults earlier, and predict peculiar events. In this paper, we survey different clustering algorithms to classify network faults and provide an evaluation based on the speed, accuracy, efficiency and cost. From accuracy point of view some of the algorithms yield high accuracy as indicated in the literature. Out of them Fuzzy Cluster-Means is considered more suitable for network faults classification based on the context in which they are applied.
|Publication status||Published - Oct 2013|
|Event||17th International Conference on Distributed Computer and Communication Networks - Moscow, Russian Federation|
Duration: 7 Oct 2013 → 10 Oct 2013
|Conference||17th International Conference on Distributed Computer and Communication Networks|
|Period||7/10/13 → 10/10/13|