Network Intrusion Detection System for Cloud Computing Security Using Deep Neural Network Framework

Abstract

Cloud computing is the only technology available in today’s era that enables the end users or the consumers to use the cloud-based services deployed by the service providers with high on-demand availability, secure, as well as cost effectiveness in nature. This computing paradigm is surrounded by several attacks that prevents the service providers from providing their services and in turn hampers the trust of the end users. This paper aims to provide a network intrusion detection system based on enhanced deep learning framework that protects the cloud environment against the network level attacks. NSL-KDD dataset is used for training the neural network. The proposed approach performs the preprocessing of the NSL-KDD dataset by utilizing recursive feature elimination as the feature selection technique and there after the optimized dataset is trained over fully connected neural network in order to construct a deep learning-based classification model. The proposed approach computes output of the hidden neurons of the fully connected neural network by ReLU activation function whereas the Sigmoid activation function is used on the output layer. Adam optimizer is used for reducing the loss in terms of actual and the expected output. The result thus obtained by the proposed NIDS clearly depicts the enhanced accuracy for attack classification and reduction in false positive rate at the same time increase in true positive rate.