Security Framework Implementation of Internet of Things Network to Detect Attacks

Abstract

Nowadays, Internet-of-Thing  associated to human's day to day life. It represents a network which includes devices that are resource-constrained. -Networks could face multiple types of attacks, therefor, security of  system is a necessary issue. To build a security framework, Machine-Learning could be a powerful tool for training models to detect attacks. The classifier would be trained by Random-Forest (RF) algorithm. Results present that the proposed framework achieve good performance on () nodes in the studied () network. The accuracy was 98%. By comparing with other algorithms such as CNN, hybrid CNN+LSTM, SVM, LSTM, Random-forest has the best performance.