Evidence Theory for Cyber-Physical Systems

Abstract : Telecommunications networks are exposed to new vulnerabilities and threats due to interdependencies and links between the cyber and physical layers. Within the cyber-physical framework, data fusion methodologies such as evidence theory are useful for analyzing threats and faults. Unfortunately, the simple analysis of threats and faults can lead to contradictory situations that cannot be resolved by classical models.Classical evidence theory extensions, such as the Dezert-Smarandache framework, are not well suited to large numbers of hypotheses due to their computational overhead. Therefore, a new approach is required to handle the complexity while minimizing the computational overhead. This paper proposes a hybrid knowledge model for evaluating the intersections among hypotheses. A hybrid frame of discernment is presented using a notional smart grid architecture that transforms the basic probability assignment values from the classical framework. Several analyses and simulations are conducted, with the goal of decreasing conflicts between two independent sources. A comparative analysis is performed using different frames of discernment and rules in order to identify the best knowledge model. Additionally, a computational time analysis is conducted.
Document type :
Conference papers
Complete list of metadatas

Cited literature [11 references]  Display  Hide  Download

https://hal.inria.fr/hal-01386757
Contributor : Hal Ifip <>
Submitted on : Monday, October 24, 2016 - 3:32:03 PM
Last modification on : Monday, October 24, 2016 - 3:40:28 PM

File

978-3-662-45355-1_7_Chapter.pd...
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Riccardo Santini, Chiara Foglietta, Stefano Panzieri. Evidence Theory for Cyber-Physical Systems. 8th International Conference on Critical Infrastructure Protection (ICCIP), Mar 2014, Arlington, United States. pp.95-109, ⟨10.1007/978-3-662-45355-1_7⟩. ⟨hal-01386757⟩

Share

Metrics

Record views

66

Files downloads

114