Illegal Intrusion Detection for In-Vehicle CAN Bus Based on Immunology Principle

The controller area network (CAN) bus has become one of the most commonly used protocols in automotive networks. Some potential attackers inject malicious data packets into CAN through external interfaces for implementing illegal operations (intrusion). Anomaly detection is a technique intrusion which can detect packs by comparing normal with incoming obtained from traffic. symmetric and stable state, will asymmetric when compromised. Considering in-vehicle network, symmetrically similar to immune system terms internal structure invasion threats. In this work, we use an method based on dendritic cell algorithm (DCA). However, existing studies suggest optimization methods improve accuracy classification algorithms, current parameters mostly relies manual tuning parameters, large workload. view above challenges, paper proposes new particle swarm (PSO) gravitational search (GSA) (PSO-GSA-DCA). PSO-GSA-DCA achieves adaptive parameter improves mixing algorithms using them optimize classifier. Additionally, DCA-based message attribute matching rules (measured information gain standard deviation data) are proposed three input signals (PAMP, DS, SS) DCA. experimental results show that our scheme significant improvement accuracy, reach 91.64%, lower time loss compared other correlation anomaly schemes. Our also enables tuning, solves problem models now rely tuning.