Distributed Denial of Service Attack Detection by Expert Systems

Abstract

The Denial of Service (DoS) attacks are the attacks that overload the system resources such as CPU, network bandwidth, memory and so on to prevent system to provide services any legitimate users. The Distributed Denial of Service (DDoS) attacks are DoS attacks that organized with several systems widely (BotNet) to shut down the servers. Many companies have developed many DDoS detector systems but as the attack patterns are getting more complex day by day, the prediction of DDoS attacks by a specific method with a reasonable cost still is a hard task. In this paper, we tried to detect DDoS attacks by expert systems that use the attack symptoms and histories. We used expert system because DDoS attacks algorithms and patterns are complicated increasingly and as a result, we need to learn the attack detector systems. Finally, we implemented our system with visual studio .net and compared the results with simulation software such as "Netica".

Keywords

• distributed denial of service attacks
• botnet
• expert system
• flow entropy
• bayesian networks
• fuzzy scale