Intrusion Detection for Future ITS: Integrated Knowledge Graph and Artificial Intelligence

The increasing connectivity and automation in the Internet of vehicles (IoV) have significantly heightened the risk of network attacks, making intrusion detection systems (IDS) a crucial component of security measures in intelligent transportation systems (ITS). To address this challenge, we propose an advanced intrusion detection method integrating knowledge graph (KG) and artificial intelligence (AI) techniques, termed IDS-IKGAI, to enhance the security of IoV infrastructures. In our proposed scheme, we first preprocess an intrusion detection dataset specific to IoV, i.e., feature selection and extraction, that can be represented as triples using the resource description framework (RDF). These RDF triples are used to construct a knowledge graph, capturing the semantic relationships among the features. Next, we map the knowledge graph to a vector space, to build a labeled dataset for machine learning. To train and predict potential intrusions, the random forest (RF) and light gradient boosting machine learning (LightGBM) algorithms are investigated. Experimental evaluations demonstrate the effectiveness of our proposed scheme, with around F1 scores of 99.99% for RF and 99.93% for LightGBM, outperforming conventional benchmark models.