Neutrosophic Set and Machine Learning Models for Detection of DoS Attack Resilience

Security has been a major problem in in-vehicle networks (VNs) in recent years, assaults that broadcast a deluge of packets, including Denial of Service (DoS) and Distributed Denial of Service (DDoS) assaults, might put the network at risk. Consequently, malicious traffic is clogging the network's resources. In this regard, the literature currently in publication has offered several strategies for dealing with DoS and DDoS attacks. In contrast to the conventional methods, this work uses machine learning (ML) to suggest an intelligent intrusion detection system (IDS). To mitigate DDoS assaults, the suggested IDS makes use of an application layer dataset that is openly accessible. Then we use the neutrosophic set model to select the best ML model under different evaluation matrices. The MABAC method is used to select the best model. A neutrosophic set is used to overcome uncertainty information. Our method's experimental validation involves a thorough assessment of various machine learning models, including naïve Bayes (NB), decision trees (DT), and random forests (RF). Surprisingly, the average system accuracy of 0.99 obtained from the combined accuracy of these models outperforms current techniques. In contrast to traditional methods, our proposed IDS is highly effective and performs well in identifying DoS and DDoS attacks in VN.