Open AccessUse of MARKOV Chain for Early Detecting DDoS Attacks
Author(s) -
YuKang Tu,
Jianming Chen
Publication year - 2021
Publication title -
international journal of network security and its applications/international journal of network security and applications
Language(s) - English
Resource type - Journals
eISSN - 0975-2307
pISSN - 0974-9330
DOI - 10.5121/ijnsa.2021.13401
Subject(s) - botnet , denial of service attack , application layer ddos attack , computer science , markov chain , computer security , machine learning , the internet , world wide web
DDoS has a variety of types of mixed attacks. Botnet attackers can chain different types of DDoS attacks to confuse cybersecurity defenders. In this article, the attack type can be represented as the state of the model. Considering the attack type, we use this model to calculate the final attack probability. The final attack probability is then converted into one prediction vector, and the incoming attacks can be detected early before IDS issues an alert. The experiment results have shown that the prediction model that can make multi-vector DDoS detection and analysis easier.