クライアントのhttpリクエスト送信動作に着目したボット検知手法
Year: 2019
Type: article
Abstract: Attack detection comes to be more important because distributed denial of service (DDoS) attacks cause serious disruption to web servers. In addition, we have to achieve a further task to tell a new phenomenon called flash events and DDoS attacks apart. In recent years, the emergence of flash events, which are spikes in network traffic from legitimate users, has been seen frequently. Flash events are often recognized as DDoS attacks by mistake because of their large amount of traffic and numerous clients. We need to react to them in a different way, though their behaviors are similar. To realize this, we must identify the factors of them. In this paper, we propose a bot detection system that focuses on the HTTP request transmission behavior of clients to identify increases in traffic and clients. In an evaluation experiment, we evaluated the detection accuracy using datasets of flash events and DDoS attacks.
Institution: University of Miyazaki
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7145652223