A4 Refereed article in a conference publication
Distributed Internal Anomaly Detection System for Internet-of-Things
Authors: Thanigaivelan NK, Nigussie E, Kanth RK, Virtanen S, Isoaho J
Editors: IEEE
Conference name: IEEE Annual Consumer Communications and Networking Conference
Publishing place: NEW YORK, NY
Publication year: 2016
Book title : 2016 13th IEEE Annual Consumer Communications & Networking Conference (CCNC)
Journal name in source: 2016 13TH IEEE ANNUAL CONSUMER COMMUNICATIONS & NETWORKING CONFERENCE (CCNC)
Series title: IEEE Consumer Communications & Networking Conference
Number in series: 13
First page : 319
Last page: 320
Number of pages: 2
ISBN: 978-1-4673-9291-4
eISBN: 978-1-4673-9292-1
ISSN: 2331-9860
DOI: https://doi.org/10.1109/CCNC.2016.7444797(external)
We present overview of a distributed internal anomaly detection system for Internet-of-things. In the detection system, each node monitors its neighbors and if abnormal behavior is detected, the monitoring node will block the packets from the abnormally behaving node at the data link layer and reports to its parent node. The reporting propagates from child to parent nodes until it reaches the root. A novel control message, distress propagation object (DPO), is devised to report the anomaly to the subsequent parents and ultimately the edge-router. The DPO message is integrated to routing protocol for low-power and lossy networks (RPL). The system has configurable profile settings and it is able to learn and differentiate the nodes' normal and suspicious activities without a need for prior knowledge. It has different subsystems and operation phases at data link and network layers, which share a common repository in a node. The system uses network fingerprinting to be aware of changes in network topology and nodes' positions without any assistance from a positioning system.
