A Context-Based Decision-Making Trust Scheme for Malicious Detection in Connected and Autonomous Vehicles

Eze, E. O., Keates, S., Pedram, K., Esfahani, A. and Odih, U. (2022) A Context-Based Decision-Making Trust Scheme for Malicious Detection in Connected and Autonomous Vehicles. In: 2022 International Conference on Computing, Electronics & Communications Engineering (iCCECE), 17-18 Aug 2022, Southend, United Kingdom. (In Press)

[thumbnail of Copyright © 2022, IEEE] Text (Copyright © 2022, IEEE)
Keates 2022.pdf - Published Version
Available under License All Rights Reserved.

Download (1MB)

Abstract

The fast-evolving Intelligent Transportation Systems (ITS) are crucial in the 21st century, promising answers to congestion and accidents that bother people worldwide. ITS applications such as Connected and Autonomous Vehicle (CAVs) update and broadcasts road incident event messages, and this requires significant data to be transmitted between vehicles for a decision to be made in real-time. However, broadcasting trusted incident messages such as accident alerts between vehicles pose a challenge for CAVs. Most of the existing-trust solutions are based on the vehicle's direct interaction base reputation and the psychological approaches to evaluate the trustworthiness of the received messages. This paper provides a scheme for improving trust in the received incident alert messages for real-time decision-making to detect malicious alerts between CAVs using direct and indirect interactions. This paper applies artificial intelligence and statistical data classification for decision-making on the received messages. The model is trained based on the US Department of Technology Safety Pilot Deployment Model (SPMD). An Autonomous Decision-making Trust Scheme (ADmTS) that incorporates a machine learning algorithm and a local trust manager for decision-making has been developed. The experiment showed that the trained model could make correct predictions such as 98% and 0.55% standard deviation accuracy in predicting false alerts on the 25% malicious data.

Publication Type: Conference or Workshop Items (Paper)
Uncontrolled Keywords: Connected vehicles, Roads, Decision making, Psychology, Predictive models, Real-time systems, Safety Context-based, decision-making, features selection and classifications, machine learning, trust
Divisions: Academic Areas > Department of Engineering, Computing and Design > Computing
Event Title: 2022 International Conference on Computing, Electronics & Communications Engineering (iCCECE)
Event Location: Southend, United Kingdom
Event Dates: 17-18 Aug 2022
Related URLs:
SWORD Depositor: Publications Router Jisc
Depositing User: Publications Router Jisc
Date Deposited: 01 Nov 2022 11:25
Last Modified: 18 Aug 2024 01:10
URI: https://eprints_test.chi.ac.uk/id/eprint/6475

Actions (login required)

View Item
View Item

Our address

I’m looking for