DDoS-FOCUS: A Distributed DoS Attacks Mitigation using Deep Learning Approach for a Secure IoT Network

The fast growth of the Internet of Things devices and communication protocols poses equal opportunities for lifestyle-boosting services and pools for cyber attacks. Usually, IoT network attackers gain access to a large number of IoT (e.g., things and fog nodes) by exploiting their vulnerabilities to set up attack armies, then attacking other devices/nodes in the IoT network. The Distributed Denial of Service (DDoS) flooding-attacks are prominent attacks on IoT. DDoS concerns security professionals due to its nature in forming sophisticated attacks that can be bandwidth-busting. DDoS can cause unplanned IoT-services outages, hence requiring prompt and efficient DDoS mitigation. In this paper, we propose a DDoS-FOCUS; a solution to mitigate DDoS attacks on fog nodes. The solution encompasses a machine learning model implanted at fog nodes to detect DDoS attackers. A hybrid deep learning model was developed using Conventional Neural Network and Bidirectional LSTM (CNN-BiLSTM) to mitigate future DDoS attacks. A preliminary test of the proposed model produced an accuracy of 99.8% in detecting DDoS attacks.

Automated summary

The fast growth of Internet of Things devices and communication protocols presents both opportunities for lifestyle-enhancing services and risks of cyber attacks. Distributed Denial of Service (DDoS) flooding attacks are a prominent threat to IoT, which require prompt mitigation due to their potential to cause unplanned outages. In this paper, a solution called DDoS-FOCUS is proposed, utilizing a machine learning model implemented at fog nodes to detect and mitigate DDoS attacks. The preliminary test showed an accuracy rate of 99.8% in detecting DDoS attacks.