Detection of DNS DDoS Attacks with Random Forest Algorithm on Spark

• DOI: 10.1016/j.procs.2018.07.177
• Publication Date: 2018-07-30T15:28:46Z
• Authors: Liguo Chen; Yuedong Zhang; Qi Zhao; Guanggang Geng; ZhiWei Yan
• Keywords: SPARK (programming language); Robustness

Domain Name System(DNS) is one of the most foundational and essential services on Internet, security robustness DNS are great significance. However, stable operation has been threatened by Distributed Denial Service(DDoS) for quite a long time, especially when number registered names of. CN over 20 million November 11, 2016. According to our observation, frequency volume-based DDoS attacks increased rapidly in recent years, attack happened, not only authoritative servers were affected, Top Level Domain(TLD) also suffered lot. In this paper, model based Random Forest[1] applied traffic classification with an accuracy 99.2% Spark. The result shows that could be used deal large-scale query flows, which fast enough practice.