Multiple Discriminant Data Analysis for Distributed Denial of Service Attacks
Safak Kayikci

Denial of Service attacks is achieved by leaks on transmission protocols and use security bugs in applications. From day by day, these attacks are continuously expanding with new difficulties on how to struggle with their influences. To preserve the system from these attacks, some basic protection procedures should be applied. In this study, four kinds of DDoS attacks (HTTP Flood, UDP Flood, Smurf, and SQL injection) data are analyzed with mixture discriminant analysis, quadratic discriminant analysis, and regularization discriminant analysis. It is observed that the packet rate is the most effective feature in attack type with packet size and the number of packets.

Full Text: PDF     DOI: 10.15640/jcsit.v8n1a1