A SECURITY FRAMEWORK FOR SYBIL ATTACKS ON VEHICULAR AD-HOC NETWORKS WITH HYBRID TRUST-BASED DETECTION

• Main Author: Rhamdhan, Agria
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/46407
• Institution: Institut Teknologi Bandung

The application of IoT in Vehicular Ad-Hoc Networks (VANET) allows the realization of intelligent transportation systems to ensure the comfort and safety of road users. However, Sybil attacks had a substantial impact on VANET. It can carry out malicious activities such as disrupting routing, causing traffic jams, bottlenecks, and even accidents. This thesis proposes an improved framework to increase the accuracy of Sybil attack detection in VANET. The solution is based on the idea that the combination of evaluation centers based on hybrid trust management and local detection based on data-centric and machine learning can increase the detection quality. The proposed fremework overcomes the problems associated with privacy-preserving, safety considerations, and real-world implementations. The performance of the proposed method is evaluated using the framework for misbehavior detection dataset to determine the effectiveness of the concept. From the simulation work, it was found that it increases the accuracy rate to 99.28% and decreases the current method false positive and false negative to 8.9% and 20%. The method proposed in this thesis is expected to be used as a framework of Sybil detection standard in the Intelligent Transportation System.