Malware Detection in the IoT Home Network

Malware Detection in the IoT Home Network

연구 분야: Networking

학회: World Congress in Computer Science, Computer Engineering & Applied Computing

The recent development of Internet of Things (IoT)-based networks, devices, and applications has led to concerns regarding the security of such technology. In particular, IoT networks set up in a home (smart homes) can be vulnerable to cyber threats due to a lack of security measures, such as secure passwords. This research proposes to secure IoT home networks from cyber threats by identifying irregularities in network protocols, a common indicator of malicious activity. Three methods are used for this task. The first is real-time monitoring of network protocols using time series analysis. The second uses network protocol data as inputs to machine learning algorithms, which are tasked with detection of malicious activity. The third approach uses an IoT-custom firewall to block access to IoT devices from irregular network traffic. The approaches are each demonstrated on network traffic datasets, including CICIoT2023. The results show the machine learning algorithms can detect malicious activity with over 95% accuracy. The custom firewall is shown to block HTTP requests. In the future it is possible to expand the real-time monitoring with more sophisticated outlier detection methods, such as autoencoders.