Identifying Anomalous Patterns of Network-wide Traffic using Flow Clustering

Raimir Holanda Aldri L. dos SantosJosé Neuman de Souza

The network traffic behavior is constantly changing due to issues such as high demand, network attacks and emergence of new services. Although network traffic characterization is a well-known task, it must mainly be effective in real-time anomalous situations in order to keep the network performance. Classical approaches have been modified in order to attempt those requirements.This work presents a methodology for identifying patterns into broadband network traffic and aims to detect traffic anomalies. This methodology is based onflow clustering analysis, a typical method employed for discovering associations and structures in collected data. The clustering analysis enables the extraction of templates of dataflows, allowing the mapping and monitoring of possible anomalies on network traffic. Furthermore, those templates represent statistically the original data.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: