A SYSTEM FOR DENIAL OF SERVICE ATTACK DETECTION BASED ON MULTIVARIATE CORRELATION ANALYSIS

Over the internet, in our day to day life we are working with interconnect systems such as web servers, database servers, cloud computing servers. These are the systems which are dealing with so many requests and responds them as legitimate requests. But the systems can be targeted by the attackers using Denial-of-Service(DoS) attack for temporary or permanent failure of the system. Denial-of-Service attack causes serious impact on the performance of the server with many a times the server gets down and stops processing the requests especially the genuine or legitimate requests. It happened so, because the server remains busy with the fake requests sent from the attackers by serving those fake requests. So, to increase the efficiency and the performance of the server, we must need to detect and avoid the DoS attacks. In this paper, we present a DoS attack detection system using features normalization and triangle area map techniques under Multivariate Correlation Analysis(MCA) which are useful for accurate traffic characterization. Traffic Characterization is done by extracting geometric correlation between network traffic features. Our DoS attack detection system can detect both known and unknown DoS attacks since it implements the principle of anomaly based detection for attack reorganization. Effectiveness of the system is increased because of its capability to learn the new patterns of legitimate network traffic. Triangle-area-based technique is used to speed up the process. Detection of SQL injection is also introduced in the system for security purpose of the stored legitimate profiles. The system designed to carry out attack detection is a question-answer portal i.e. a web