Framework for Detection of Malicious Activities in IoT Networks using Keras Deep Learning Library

Abstract

Secure and reliable services of the smart cities are generally depend on the reliable services provided by different devices of the Internet of Things ecosystem and Internet of Things backbone networks. In order to provide secure and reliable services, there is need to install intrusion detection mechanism to detect malicious and intrusions activities of the malicious attackers on the IoT network. This paper presents a framework for detection of malicious activities in IoT Backbone Networks using Keras Deep Learning Library. The proposed framework uses four different deep learning models such as Multi Layer Perceptron (MLP), Convolutional Neural Networks (CNN), Deep Neural Networks (DNN) and Autoencoder for predicting the malicious attacks. Performance evaluation of the proposed framework is done using two well known datasets such as UNSW-NB15 and NSL-KDD99 and its result analysis are discussed in terms of accuracy, RMSE and F1-score.