Evaluating Granularity in Markov Chain-Based Trust Models for Vehicular Ad Hoc Networks (VANETs)

TL;DR

This paper compares different Markov chain models with varying granularity to evaluate their effectiveness in capturing driver behaviour for trust management in VANETs, demonstrating that higher granularity improves behavioural representation.

Contribution

It introduces and evaluates three Markov chain models with increasing states to analyze driver trust dynamics in VANETs, highlighting the benefits of higher granularity.

Findings

Higher state models better capture complex driver behaviours

Granular models improve trust assessment accuracy

Increased states enhance security in VANETs

Abstract

Trust management is a critical research pillar in Vehicular Ad Hoc Networks (VANETs), where the reliability of shared data depends entirely on driver integrity. In these networks, a driver's reputation is dynamically constructed based on the veracity of their recent message history: consistent reliability builds trust, while frequent misinformation leads to exclusion. This study analyses driver announcement characteristics by modelling behavioural transitions-specifically the frequency and nature of shifts between "good" and "bad" states. To facilitate this analysis, three distinct Markov chain-based behavioural models are evaluated with varying degrees of granularity: a 4-state model, a 7-state model, and a high-resolution 11-state model. By simulating announcement and reporting patterns, each model's ability to reflect nuanced behavioural shifts is assessed. Our results confirm that increasing the number of trust states significantly enhances the system's ability to capture complex, dynamic driver behaviours, providing a more robust framework for security in VANETs.