IPv6 flood attack detection based on epsilon greedy optimized Q learning in single board computer

<span lang="EN-US">Internet of things is a technology that allows communication between devices within network. Since this depends on network to communicate, the vulnerability exposed increased significantly. Furthermore, use internet protocol version 6 (IPv6) as successor 4 (IPv4) constituted significant problem for Hence, was exploitable flooding attacks in IPv6 As countermeasure against flood, study designed an flood attack detection by using epsilon greedy optimized Q learning algorithm. According evaluation, agent with 0.1 could reach 98% accuracy and 11,550 rewards compared other agents. When control models, also most accurate algorithms followed neural (NN), K-nearest neighbors (KNN), decision tree (DT), naive Bayes (NB), support vector machine (SVM). Besides that, used more than 99% single central processing unit (CPU). will not hinder (IoT) multiple processors. Thus, we concluded proposed has high feasibility board computer (SBC).</span>