Intrusion Detection System Using GWO-OptimizedLogistic Regression

The Internet of Things can grow and disseminate with newly developed technologies. These devices havelimited resources, which can be exploited to generate distributed denial-of-service attacks that are widelydistributed and extended until the server is completely reduced or stopped. Within the scope of this research,we suggest a framework for detecting distributed denial-of-service attacks that ion fog computing. Theproposed Gray Wolf Optimization Logistic Regression (GWO-LR) system comprises an algorithm for logisticRegression that is trained with the help of an algorithm for Gray Wolf Optimization GWO. The GWO-LRsolves the classification problem in the UNSW Bot-IoT ۲۰۱۸ database. The results showed that the classifiercould detect attacks with a high accuracy of ۹۸.۸۸% and an F-measure of ۹۹%.

Keywords:

IoT , Gray Wolf Optimization , Logistic Regression , Intrusion Detection