دراسة تأثير المجال الأعظمي المعقول للتغطية على أداء خوارزميات الكشف عن سوء السلوك في شبكات VANET

Abstract

The maximum plausible coverage range in Vehicular Ad Hoc Networks (VANETs) is the distance within which vehicles can exchange beacon messages to warn drivers about road conditions. Although extending this range allows receiving messages from multiple vehicles, it does not guarantee the reliability of these messages, which may open the door to malicious attacks that affect driving safety. The sudden appearance range represents a situation where a vehicle suddenly appears within the range of another vehicle without a gradual exchange of messages, increasing the chances of attacks or interference.

To address these challenges, we propose a trust-based algorithm relying on a weighted average for inspection (TrustWeightCheck), which integrates trust with dynamic weights for checks, such that the weight of a failed check is increased while the weight of a successful check is reduced without neglecting any result. The algorithm’s performance was evaluated under two coverage ranges: 420 meters and 600 meters. The results showed that increasing the range slightly affects detection performance, with a minor increase in detection time and a limited decrease in F1-Score metrics. However, the proposed algorithm outperformed others in detection speed, achieving an average detection time of 0.13 seconds along with a high F1-Score rate of 58%. Despite higher entropy in identifying the attacking vehicle due to pseudonym changes that reduce traceability, the algorithm demonstrated a good balance by relying on all checks compared to algorithms based on a single criterion, neighbor trust, or preliminary attack type predictions.