Network traffic features for anomaly detection in specific industrial control system network

Matti Mantere (Corresponding Author), Mirko Sailio, Sami Noponen

Research output: Contribution to journalArticleScientificpeer-review

23 Citations (Scopus)

Abstract

The deterministic and restricted nature of industrial control system networks sets them apart from more open networks, such as local area networks in office environments. This improves the usability of network security, monitoring approaches that would be less feasible in more open environments. One of such approaches is machine learning based anomaly detection. Without proper customization for the special requirements of the industrial control system network environment, many existing anomaly or misuse detection systems will perform sub-optimally. A machine learning based approach could reduce the amount of manual customization required for different industrial control system networks. In this paper we analyze a possible set of features to be used in a machine learning based anomaly detection system in the real world industrial control system network environment under investigation. The network under investigation is represented by architectural drawing and results derived from network trace analysis. The network trace is captured from a live running industrial process control network and includes both control data and the data flowing between the control network and the office network. We limit the investigation to the IP traffic in the traces.
Original languageEnglish
Pages (from-to)460-473
Number of pages13
JournalFuture Internet
Volume5
Issue number4
DOIs
Publication statusPublished - 2013
MoE publication typeA1 Journal article-refereed

Fingerprint

Control systems
Learning systems
Trace analysis
Network security
Local area networks
Process control
Monitoring

Keywords

  • industrial control systems
  • anomaly detection
  • machine learning
  • network security

Cite this

@article{d9e701e7145d4124a628120a6ab2f2f1,
title = "Network traffic features for anomaly detection in specific industrial control system network",
abstract = "The deterministic and restricted nature of industrial control system networks sets them apart from more open networks, such as local area networks in office environments. This improves the usability of network security, monitoring approaches that would be less feasible in more open environments. One of such approaches is machine learning based anomaly detection. Without proper customization for the special requirements of the industrial control system network environment, many existing anomaly or misuse detection systems will perform sub-optimally. A machine learning based approach could reduce the amount of manual customization required for different industrial control system networks. In this paper we analyze a possible set of features to be used in a machine learning based anomaly detection system in the real world industrial control system network environment under investigation. The network under investigation is represented by architectural drawing and results derived from network trace analysis. The network trace is captured from a live running industrial process control network and includes both control data and the data flowing between the control network and the office network. We limit the investigation to the IP traffic in the traces.",
keywords = "industrial control systems, anomaly detection, machine learning, network security",
author = "Matti Mantere and Mirko Sailio and Sami Noponen",
year = "2013",
doi = "10.3390/fi5040460",
language = "English",
volume = "5",
pages = "460--473",
journal = "Future Internet",
issn = "1999-5903",
publisher = "MDPI",
number = "4",

}

Network traffic features for anomaly detection in specific industrial control system network. / Mantere, Matti (Corresponding Author); Sailio, Mirko; Noponen, Sami.

In: Future Internet, Vol. 5, No. 4, 2013, p. 460-473.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Network traffic features for anomaly detection in specific industrial control system network

AU - Mantere, Matti

AU - Sailio, Mirko

AU - Noponen, Sami

PY - 2013

Y1 - 2013

N2 - The deterministic and restricted nature of industrial control system networks sets them apart from more open networks, such as local area networks in office environments. This improves the usability of network security, monitoring approaches that would be less feasible in more open environments. One of such approaches is machine learning based anomaly detection. Without proper customization for the special requirements of the industrial control system network environment, many existing anomaly or misuse detection systems will perform sub-optimally. A machine learning based approach could reduce the amount of manual customization required for different industrial control system networks. In this paper we analyze a possible set of features to be used in a machine learning based anomaly detection system in the real world industrial control system network environment under investigation. The network under investigation is represented by architectural drawing and results derived from network trace analysis. The network trace is captured from a live running industrial process control network and includes both control data and the data flowing between the control network and the office network. We limit the investigation to the IP traffic in the traces.

AB - The deterministic and restricted nature of industrial control system networks sets them apart from more open networks, such as local area networks in office environments. This improves the usability of network security, monitoring approaches that would be less feasible in more open environments. One of such approaches is machine learning based anomaly detection. Without proper customization for the special requirements of the industrial control system network environment, many existing anomaly or misuse detection systems will perform sub-optimally. A machine learning based approach could reduce the amount of manual customization required for different industrial control system networks. In this paper we analyze a possible set of features to be used in a machine learning based anomaly detection system in the real world industrial control system network environment under investigation. The network under investigation is represented by architectural drawing and results derived from network trace analysis. The network trace is captured from a live running industrial process control network and includes both control data and the data flowing between the control network and the office network. We limit the investigation to the IP traffic in the traces.

KW - industrial control systems

KW - anomaly detection

KW - machine learning

KW - network security

U2 - 10.3390/fi5040460

DO - 10.3390/fi5040460

M3 - Article

VL - 5

SP - 460

EP - 473

JO - Future Internet

JF - Future Internet

SN - 1999-5903

IS - 4

ER -