Mitigation of attacks via improved network security in IOT network environment using RNN

Abstract

Internet of Things (IoT) has become one of the emerging communication paradigms in recent years. Distributed denial of service (DDoS) attacks is becoming more common in IoT implying a growing need for security and authentication. Current advancements in IoT indicate fundamental changes in global communication infrastructures. Many facets of urban lives in smart cities are significantly impacted by increased interoperability of smart communication technologies. This paper proposes an IoT network-based threat mitigation strategy based on Recurrent Neural Network (RNN) algorithm. Using pre-processed and feature-extracted data, RNN seeks to categorise attributes associated with attacks. XBoost model selects features after the datasets have been preprocessed using min-max scaling technique. The simulations are executed over KDD datasets and assessed for the metrics of accuracy, precision, recall, and f-measure. The findings of this work demonstrated that in both training and testing datasets, the proposed RNN based schema achieves high degrees of classification accuracy.

Keywords

• RNN
• Attack mitigation
• IoT
• KDD datasets
• Deep learning