A machine learning model for detection of man in the middle attack over unsecured devices

• Main Authors: Mantoo, Bilal Ahmad, Kaur, Parveen
• Format: Conference Proceeding
• Language: English
• Subjects: Computer networks; Cybersecurity; Internet of Things; Machine learning; Networks
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0109151

Due to increased internet accessibility, Internet of Things (IoT) devices grows exponentially and almost everything is got connected with it. But at the same time the security and privacy of IoT is of more concern .When you place more gadgets in the market their security is of equally important. The attacks are proving huge loss to networks and at the same time can lead to breach of our security. In comparison to previous established computer networks; IoT networks have improved features which make the revealing of attacks over such gadgets a challenging task. The different mix of technology like software, protocols, hardware exposes them to loopholes. These loopholes are persisting in the network and obfuscate the benign traffic. The attackers are so canny that they dynamically change their behavior in order of being detected. In order to adapt a proper pattern to identify the features of IoT attacks, like MITM (man in the middle attack), a Machine learning based model for detection of such attacks is being proposed. In this paper, crucial features of IoT attacks like man in the middle attack (MITM) are being analysed and the a Machine learning model is being developed to detect this attack efficiently using K Neighrest Neighbor (KNN). The model uses different sets of data obtained using wireshark and the results obtained shows the accuracy of 0.98