An Online Entropy-based DDoS Flooding Attack Detection System with Dynamic Threshold
Loïc D. Tsobdjou,Samuel Pierre,Alejandro Quintero,Loic D. Tsobdjou
DOI: https://doi.org/10.1109/tnsm.2022.3142254
2022-01-01
IEEE Transactions on Network and Service Management
Abstract:Distributed denial of service attacks are cyber-attacks that target the availability of servers. As a result, legitimate users no longer have access to the service. This can have a negative impact on an organization, such as lack of reputation and economic losses. Therefore, it is important to design defense mechanisms against these attacks. There are systems for detecting distributed denial of service attacks in the literature, which still have various shortcomings. Some of these systems detect the presence of attack traffic without identifying the attack packets or flows. Others use static thresholds and therefore cannot adapt to changes in legitimate traffic. In this paper, we propose an online system that aims to detect flooding attacks in a short timeframe and a client–server environment. The proposed detection system consists of five modules, namely features extraction and connections construction, suspicious activity detection, attack connections detection, alert generation and threshold update. The suspicious activity detection module calculates the normalized Shannon entropy by considering the source Internet Protocol address as a random variable. Suspicious activity is detected when the computed entropy is below a threshold. The threshold calculation is based on Chebyshev’s theorem. We propose a dynamic threshold algorithm to track changes in legitimate traffic. We evaluate the proposed system through simulations and using a publicly available dataset. Compared to other similar works, the proposed detection system has a better performance in terms of detection rate, false positive rate, precision and overall accuracy.
computer science, information systems