خوارزمية الثقة المعتمدة على المتوسط الموزون للتحديد سلوك العربات في VANET

Abstract

VANET networks represent a vital solution for enhancing driver safety by raising awareness about road landmarks and guiding drivers to adhere to them. However, these networks can pose a security threat when exploited for malicious purposes, such as eavesdropping on drivers, tracking them, or manipulating and retransmitting messages, potentially leading to severe accidents, including fatalities. To counter these threats, the study proposes a comprehensive analysis of misbehavior detection algorithms in VANET networks. This involves rigorous testing to validate the exchanged messages within the network. A weighted average-based algorithm is proposed to calculate trust in determining whether a vehicle's behavior is legitimate or not. The proposed algorithm was evaluated under scenarios of high and low attack density, as well as under high and normal traffic density. The results showed that the Precision was high at 51%, and the F1-score was 50% when applied in high traffic density with a high attacker density.