Anomaly-Based Detection Approach To Detect The Flash Crowd Attack During The Flash Event

• Main Author: Al-Saleem, Samer Abdulsada Mutlag
• Format: Thesis
• Language: English
• Published: 2017
• Subjects: TK5105 Computer networks and Data transmission systems
• Online Access: http://eprints.usm.my/47734/1/ANOMALY-BASED%20DETECTION%20APPROACH%20TO.pdf%20cut.pdf; http://eprints.usm.my/47734/

The goal of the thesis is to propose an approach that can detect flash crowd attacks that target web servers during flash events. The proposed approach consists of four phases to achieve the goal of the research. The proposed approach is evaluated using three different datasets distributed in three scenarios. The first scenario is aimed to validate the detection accuracy of the proposed approach when a web server is in a normal situation (when there is no threat targeting the server). The second scenario (flash event dataset) is aimed to improve the accuracy of the proposed approach in detecting anomalous behaviour when the server is facing a flash event. The third scenario (flash crowd attack dataset) is aimed to verify and validate the accuracy of the proposed approach in terms of detecting the flash crowd attack that targets the web server during the flash event.