Academic Journals Database
Disseminating quality controlled scientific knowledge

Clustering of High-Volume Data Streams In Network Traffic

Author(s): M. Vijayakumar | R.M.S. Parvathi

Journal: International Journal of Computer Science and Information Security
ISSN 1947-5500

Volume: 8;
Issue: 6;
Start page: 229;
Date: 2010;
Original page

Keywords: Traffic analysis | network management | clustering | frequent Item set | hierarchical clustering.

The thesis concerned with the problem of mining network traffic data discovering useful associations, relationships, and groupings in large collections of data. Mathematical transformation algorithms have proven effective at reducing the content of multilingual, unstructured data into a vector that describes the content. Such methods are particularly desirable in fields undergoing information explosions, such as network traffic analysis, bio-informatics, and the intelligence community. In response, traffic mining methodology is being extended to improve performance and sufficiently scalable.The usage of data flow collected from site routers for various analysis i.e., network performance characterization, investigating computer security incidents and their prevention, network traffic statistics, and others. Currently, the data flow analysis is built as a distributed system to collect data from multiple routers, both at the edge of the site network as well as from local routers and multilayer switches. Average per day volume is about 2GBytes of raw data. Despite a high volume of collected information, some analysis is conducted in near real time to satisfy demands of users communities for quick results.The proposed work present an efficient clustering means to analyze experimental results for traffic data streams nature (symmetric and asymmetric). As summary, this paper describe a system designed to satisfy three primary goals i.e., real-time concept mining of high-volume data streams, dynamic data flow into a relational hierarchy; and adaptive reorganization of the traffic data hierarchy in response to evolving circumstances and network traffic time to time. The proposed clustering network traffic data flow collection and analysis system describe traffic characterization and network performance estimation for the data flow centre. The system checks the traffic consistency for End To End circuits and Policy Based Routing and finally, profiling of host's traffic to keep track of their typical behavior to prevent accidental blocking by site IDS system.
Affiliate Program     

Tango Rapperswil
Tango Rapperswil