Game Theory based Modified Naïve-bayes Algorithm to detect DoS attacks using Honeypot
Rajesh Shrivastava, Saradhi Ramakrishna, Chittaranjan Hota · 2019
The omnipresent nature of the Internet makes everything vulnerable against DoS attacks in IoT environments. An adversary tries to mount attacks to serve his/her nefarious purposes. Preventing IoT network from this type of attack is a complex and challenging problem. A game theory model is best suited to explore the network security problems. Attack and defense is a continuous game in network security. A defender tries to neutralise an attack by using his secret defensive method while an attacker attempts to defeat the security mechanism. They both compete against each other's methods. In this paper, we propose a game-theoretic model which uses the nash equilibrium to expose an adversary's attack attempt and verify the defense mechanism against the same attack. In this work, a combination of Naïve- bayes with K-means clustering algorithm is used to evaluate the unlabelled data and explore malware attacks. We have also proposed a solution to mitigate IoT attacks using a firewall and an IDS.