TBNDG: Trust-Based Malicious Node Detection for Grey-Hole Attacks in VANET’s

Vehicular advertising networks (VANETs) are vulnerable to various security threats due to their dynamic and decentralized nature. One of the most critical threats is the grey hole attack, where malicious nodes selectively drop packets, which poses a significant threat to communication reliability and security in VANETs. This paper proposes a robust approach to detect malicious nodes involved in grey-hole attacks. Our method uses the concept of trust to distinguish between trusted and untrusted nodes in the network. Each node maintains a trust score for neighbouring nodes based on their behaviour and influence. Trust is calculated through direct observation, such as packet forwarding and message integrity checks, and indirect feedback from other nodes in the network. To effectively detect grey-hole attackers, we use a multi-dimensional trust evaluation mechanism that considers several factors, such as packet delivery consistency, message integrity, and node behaviour history. By summing up these factors, we calculate a comprehensive confidence score for each point that allows us to identify potential malicious actors. In addition, we introduced updates to the trust and consensus mechanisms between neighbouring nodes to improve the strength of our approach against reconciliation and false accusations. Nodes periodically exchange trust information and check the consistency of their trust estimates, thus reducing the impact of false positives and negatives. We evaluate the performance of our proposed trust-based detection scheme through extensive simulations using real-world VANET scenarios. The results show that it detects grey-hole attackers with high accuracy, low computational complexity, and effective communication. In summary, our proposed model TBNDG calculates the performance in terms of PDR, End-To-End Delay, Routing Overhead, Packet Loss Ratio and Throughput than the existing model