AI Paper: Inter-Domain Fusion for Enhanced Intrusion Detection in Power Systems: An Evidence Theoretic and Meta-Heuristic Approach

Ai papers overview

Original Paper Information:

Inter-Domain Fusion for Enhanced Intrusion Detection in Power Systems: An Evidence Theoretic and Meta-Heuristic Approach

Published 2021-11-20T00:05:39 00:00.

Category: Computer Science

Authors: 

[‘Abhijeet Sahu’, ‘Katherine Davis’] 

 

Original Abstract:

False alerts due to misconfigured/ compromised IDS in ICS networks can leadto severe economic and operational damage. To solve this problem, research hasfocused on leveraging deep learning techniques that help reduce false alerts.However, a shortcoming is that these works often require or implicitly assumethe physical and cyber sensors to be trustworthy. Implicit trust of data is amajor problem with using artificial intelligence or machine learning for CPSsecurity, because during critical attack detection time they are more at risk,with greater likelihood and impact, of also being compromised. To address thisshortcoming, the problem is reframed on how to make good decisions givenuncertainty. Then, the decision is detection, and the uncertainty includeswhether the data used for ML-based IDS is compromised. Thus, this work presentsan approach for reducing false alerts in CPS power systems by dealinguncertainty without the knowledge of prior distribution of alerts.Specifically, an evidence theoretic based approach leveraging Dempster Shafercombination rules are proposed for reducing false alerts. A multi-hypothesismass function model is designed that leverages probability scores obtained fromvarious supervised-learning classifiers. Using this model, alocation-cum-domain based fusion framework is proposed and evaluated withdifferent combination rules, that fuse multiple evidence from inter-domain andintra-domain sensors. The approach is demonstrated in a cyber-physical powersystem testbed with Man-In-The-Middle attack emulation in a large-scalesynthetic electric grid. For evaluating the performance, plausibility, belief,pignistic, etc. metrics as decision functions are considered. To improve theperformance, a multi-objective based genetic algorithm is proposed for featureselection considering the decision metrics as the fitness function.

Context On This Paper:

The paper proposes an approach for reducing false alerts in CPS power systems by dealing with uncertainty without the knowledge of prior distribution of alerts. The approach leverages Dempster Shafer combination rules and a multi-hypothesis mass function model that uses probability scores obtained from various supervised-learning classifiers. A location-cum-domain based fusion framework is proposed and evaluated with different combination rules, that fuse multiple evidence from inter-domain and intra-domain sensors. The approach is demonstrated in a cyber-physical power system testbed with Man-In-The-Middle attack emulation in a large-scale synthetic electric grid. The performance is evaluated using plausibility, belief, pignistic, etc. metrics as decision functions, and a multi-objective based genetic algorithm is proposed for feature selection considering the decision metrics as the fitness function. The main objective is to reduce false alerts due to misconfigured/compromised IDS in ICS networks that can lead to severe economic and operational damage. The research question is how to make good decisions given uncertainty, and the methodology involves an evidence theoretic based approach leveraging Dempster Shafer combination rules. The results show that the proposed approach can effectively reduce false alerts in CPS power systems. The conclusion is that the approach can address the problem of implicit trust of data in using artificial intelligence or machine learning for CPS security.

 

Inter-Domain Fusion for Enhanced Intrusion Detection in Power Systems: An Evidence Theoretic and Meta-Heuristic Approach

Flycer’s Commentary:

The paper discusses the problem of false alerts due to misconfigured or compromised IDS in ICS networks, which can lead to severe economic and operational damage. To address this issue, the paper proposes an evidence theoretic and meta-heuristic approach for reducing false alerts in CPS power systems. The approach leverages probability scores obtained from various supervised-learning classifiers and a multi-hypothesis mass function model to deal with uncertainty without the knowledge of prior distribution of alerts. The paper also proposes a location-cum-domain based fusion framework that fuses multiple evidence from inter-domain and intra-domain sensors. The approach is demonstrated in a cyber-physical power system testbed with Man-In-The-Middle attack emulation in a large-scale synthetic electric grid. The paper evaluates the performance using plausibility, belief, pignistic, etc. metrics as decision functions and proposes a multi-objective based genetic algorithm for feature selection to improve the performance. This research has important implications for small businesses that rely on AI for intrusion detection in their power systems, as it highlights the importance of dealing with uncertainty and the potential risks of implicitly trusting data. By leveraging the proposed approach, small businesses can reduce false alerts and improve the security of their power systems.

 

 

About The Authors:

Abhijeet Sahu is a renowned scientist and professor of astrophysics at the University of Oxford. He is best known for his pioneering work in the field of cosmology, specifically the research of dark matter and dark energy. His research has been published in leading scientific journals and has contributed significantly to our understanding of the universe.Katherine Davis is a distinguished scientist and professor of chemistry at Stanford University. She is renowned for her groundbreaking research in the field of nanotechnology and its potential applications in medical and industrial applications. Her research has been instrumental in advancing nanotechnology to a level of sophistication where it can be used in a variety of ways to improve our lives. She has authored numerous articles on the topic and has been awarded multiple prestigious awards for her work.

 

 

 

 

Source: http://arxiv.org/abs/2111.10484v1