Back to search

Georgia Institute of Technology

Network based fingerprinting techniques for industrial control systems

Abstract

dc:description.abstract

Fingerprinting techniques operating over the network are proposed to identify various aspects of industrial control systems (ICSs) including software, hardware, and physical devices. First, a detailed traffic characterization is performed on several power substation networks to guide the development of the techniques. Round trip times for the resource-starved embedded devices were observed to be heavily clustered based on device type no matter how large the physical distance between them, suggesting they were largely based on processing time. This insight led to the development of cross-layer response time fingerprinting to passively identify device types based on the processing time between TCP level acknowledgments and application layer responses, with classification accuracy reaching 99% on real-world substation traffic. Complementing these techniques by addressing a different aspect of ICS networks, methods are developed to fingerprint the physical devices of the ICS. Previous work on physical fingerprinting is extended to improve relay classification from 92% to 100% and extend the scope of the methods to valves, motors, and pumps. Building on the idea behind the cross-layer response time methods, techniques are explored that expand the scope to general programmable logic controllers by generating program fingerprints from the execution times of control programs. The security of this technique is enhanced by the addition of proof-of-work functions to provide an upper bound guarantee that no additional instructions are being executed in the program. Performance of all the fingerprinting techniques are discussed with respect to their potential to contribute to a holistic, ICS-specific intrusion detection system.

Degree

thesis:*
Level thesis:degree_level
Doctoral
Department dc:contributor.department
Electrical and Computer Engineering
Grantor dc:publisher
Georgia Institute of Technology
Year dc:date.issued
2017

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Formby, David
Advisor dc:contributor.advisor
  • Beyah, Raheem A.
Committee members dc:contributor.committeemember
  • Copeland, John
  • Owen, Henry
  • Chang, Yusun
  • Zajic, Alenka
  • Zonouz, Saman

Subjects

dc:subject × 3

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/1853/60668

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Formby, David. Network based fingerprinting techniques for industrial control systems. Doctoral thesis, Georgia Institute of Technology, 2017. http://hdl.handle.net/1853/60668