The Internet of Things (IoT) paradigm is changing the way people live and work in society. The advancements in various information and communication technologies has paved the way for new possibilities and opportunities both in households and industries to build such an Internet of connected devices. As billions of IoT devices are projected to be connected to Internet by 2020, there is an urgent need for building solutions that protect these devices from misuse. Middleware based IoT application protocols play a crucial role in enabling bidirectional communication and remote controllability of IoT devices. Among the various IoT application protocols, Message Queuing Telemetry Protocol (MQTT) and Extensible Messaging and Presence Protocol (XMPP) are being widely adopted for building the IoT applications. Hence, this work aims to study the cyberattacks targeting IoT environments deployed using these application protocols. Through this thesis, a honeypot data analysis framework will be proposed to accurately detect existing and zero-day attacks targeting the MQTT and XMPP protocols. The proposed framework will leverage the benefits of proven unsupervised machine learning techniques to accurately detect attacks. The framework will also yield results in an easy to understand format which will help analysts in building effective IoT protection systems.
Please leave a comment about your rating so we can better understand how we might improve the page.