Bayesian Inference Based Decision Reliability Under Imperfect Monitoring

Abstract

Reliability of a cooperative decision mechanism is critical for the proper and accurate functioning of a networked decision system. However, adversaries may choose to compromise the inputs from different sets of components that comprise the system. Often times, the monitoring mechanisms fail to accurately detect compromised inputs; hence cannot categorize all inputs into polarized decisions: compromised or not compromised. In this paper, we propose a Bayesian inference model based on multinomial evidence to quantify reliability for a cooperative decision process as a function of beliefs associated with observations from the imperfect monitoring mechanism. We propose two reliability models: an optimistic one for a normal system and a conservative one for a mission critical system. We also provide an entropy measure that reflects the certainty or uncertainty on the calculated reliability of the decision process. Through simulation, we show how the reliability and its corresponding entropy changes as the accuracy of the underlying monitoring mechanism improves.1

Publication Date

6-29-2015

Publication Title

Proceedings of the 2015 IFIP/IEEE International Symposium on Integrated Network Management, IM 2015

Number of Pages

1333-1338

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1109/INM.2015.7140491

Socpus ID

84942627158 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84942627158

This document is currently not available here.

Share

COinS