A new procedure for misbehavior detection in vehicular ad-hoc networks using machine learning

Misbehavior detection in vehicular ad hoc networks (VANETs) is performed to improve the traffic safety and driving accuracy. All nodes VANETs communicate each other through message logs. Malicious can cause inevitable situation by sending logs with tampered values. In this work, various machine learning algorithms are used detect primarily five types of attacks namely, constant attack, offset random eventual attack. Firstly, attack detected different using binary classification. Then, new procedure created do multi classification on best chosen algorithm from techniques. The highest accuracy case obtained Naïve Bayes (100%), decision tree forest (100%) type1 type2 (98.03%, 95.56%, 95.55%) Type4, Type8 Type16 respectively. for multi-classification, (97.62%) technique. For VeReMi dataset (a public repository malicious node VANETs) used.