applied sciences Article Anomaly Detection in IoT Communication Network Based on Spectral Analysis and Hurst Exponent Paweł Dymora * and Mirosław Mazurek Faculty of Electrical and Computer Engineering, Rzeszów University of Technology, al. Powsta´nców Warszawy 12, 35-959 Rzeszów, Poland;
[email protected] * Correspondence:
[email protected] Received: 14 November 2019; Accepted: 3 December 2019; Published: 6 December 2019 Abstract: Internet traffic monitoring is a crucial task for the security and reliability of communication networks and Internet of Things (IoT) infrastructure. This description of the traffic statistics is used to detect traffic anomalies. Nowadays, intruders and cybercriminals use different techniques to bypass existing intrusion detection systems based on signature detection and anomalies. In order to more effectively detect new attacks, a model of anomaly detection using the Hurst exponent vector and the multifractal spectrum is proposed. It is shown that a multifractal analysis shows a sensitivity to any deviation of network traffic properties resulting from anomalies. Proposed traffic analysis methods can be ideal for protecting critical data and maintaining the continuity of internet services, including the IoT. Keywords: IoT; anomaly detection; Hurst exponent; multifractal spectrums; TCP/IP; computer network traffic; communication security; Industry 4.0 1. Introduction Network traffic modeling and analysis is a crucial issue to ensure the proper performance of the ICT system (Information and Communication Technologies), including the IoT (Internet of Things) and Industry 4.0. This concept defines the connection of many physical objects with each other and with internet resources via an extensive computer network. The IoT approach covers not only communication devices but also computers, telephones, tablets, and sensors used in the industry, transportation, etc.