Application of Gromov-Wasserstein Matching for Vehicular Social Network Security Profiling

Publication Date

1-1-2025

Document Type

Article

Publication Title

IEEE Transactions on Vehicular Technology

DOI

10.1109/TVT.2025.3564560

Abstract

The Vehicular Social Network (VSN) seamlessly combines social networks with the Internet of Vehicles (IoVs), fostering social connections not only among the drivers but also among the vehicles themselves. VSN security profiling helps in comprehending and ranking VSN security risks according to a range of criteria. The existing approaches to VSN security profiling and risk quantification have notable limitations. Firstly, they often overlook the intensity of interactions between VSN devices. Secondly, these methods require re-evaluation of security scores and the likelihood of attacks whenever VSN configurations change. Additionally, most current risk quantification methods cannot concurrently assess risks across multiple networks, further limiting their utility. Given these challenges, this work focuses on VSN security profiling quantitatively based on users accounts because various types of accounts are used for security control mechanisms. To address these issues, we first propose Distributed Accounts Association Algorithms (D-AAA) for matching service accounts with malicious accounts based on Gromov-Wasserstein (GW) metric, and apply accounts matching results to sort service accounts by vulnerability and compute the average risk for multiple service VSN networks. To verify the proposed D-AAA, the performance of the proposed D-AAA is evaluated through numerical experiments, including an assessment of computational complexities, association correctness, and risk evaluations. Additionally, the D-AAA are compared with other graph matching methods in terms of association correctness and runtime with real VSN datasets.

Keywords

distributed algorithms, graph matching, Gromov-Wasserstein metric, Internet of Vehicles (IoV), security profiling, Vehicular social network (VSN)

Department

Applied Data Science

Link to Full Text

Share

COinS