Characterizing the Impacts of Application Layer DDoS Attacks
Muhui Jiang, Chenxu Wang, Xiapu Luo, MiuTung Miu, Ting Chen
Abstract
Muhui Jiang, Chenxu Wang, Xiapu Luo, MiuTung Miu, Ting Chen
Abstract
Distributed Denial of Service (DDoS) attacks are still among the most urgent threats to the modern Internet. Recently, application layer DDoS attacks against web servers are becoming popular, resulting in great revenue losses to victims. A systematic evaluation on the impacts of different DDoS attack methods is vital for the protection of web servers. In this paper, we examine the impacts of application layer DDoS attacks, including existing attacks against HTTP/1.1 and the new attacks proposed by us against HTTP/2.0. Moreover, to better understand attackers' capabilities of launching severe application layer DDoS attacks, we design a new measurement method to remotely infer the performance of web servers and a method to differentiate dynamic and static URLs. We have collected and tailored 4 existing tools to launch 5 different DDoS attacks against HTTP/1.1 and developed a new DDoS tool to perform 5 different DDoS attacks against HTTP/2.0. By conducting extensive experiments in a testbed with two e-commercial websites running Apache and Nginx, we carefully evaluate the impacts of different DDoS attacks. The results show that the new remote measurement method is able to detect the effects caused by different DDoS attacks. Moreover, the attack impacts are affected by URLs, server architectures, and attack methods.
OpenAlex reports 21 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Distributed Denial of Service (DDoS) attacks are still among the most urgent threats to the modern Internet. Recently, application layer DDoS attacks against web servers are becoming popular, resulting in great revenue losses to victims. A systematic evaluation on the impacts of different DDoS attack methods is vital for the protection of web servers. In this paper, we examine the impacts of application layer DDoS attacks, including existing attacks against HTTP/1.1 and the new attacks proposed by us against HTTP/2.0. Moreover, to better understand attackers' capabilities of launching severe application layer DDoS attacks, we design a new measurement method to remotely infer the performance of web servers and a method to differentiate dynamic and static URLs. We have collected and tailored 4 existing tools to launch 5 different DDoS attacks against HTTP/1.1 and developed a new DDoS tool to perform 5 different DDoS attacks against HTTP/2.0. By conducting extensive experiments in a testbed with two e-commercial websites running Apache and Nginx, we carefully evaluate the impacts of different DDoS attacks. The results show that the new remote measurement method is able to detect the effects caused by different DDoS attacks. Moreover, the attack impacts are affected by URLs, server architectures, and attack methods.
Key concepts: Denial-of-service attack, Application layer DDoS attack, Trinoo, Testbed, Computer science, Server, Web server, Application layer