Graph Neural Network for Digital Twin-enabled Intelligent Transportation System Reliability

Digital Twin (DT) is a promising technology for executing real-time monitoring and diagnostic operations in Intelligent Transportation Systems (ITSs). Guaranteeing safety and functionality in DT-enabled ITS domain requires reliable data streams among vehicular edge and cloud network resources. Therefore, in this paper, we propose Graph Neural Network based Anomaly Detection (GNN-AD) as a generic end-to-end framework for achieving reliability in multi-dimensional data streams of ITS platforms. Leveraging unsupervised and supervised machine learning algorithms, namely BiDirectional Generative Adversarial Network (BiGAN), affinity propagation and Graph Convolutional Network (GCN), we semantically model the complex distributions of ITS data streams and perform anomaly detection utilizing graph representations. Employing DT applications and diverse sensory measurements from the BeIntelli smart mobility platform of TU Berlin, we compare the performance of our approach with traditional and deep learning based methods in terms of detection accuracy and Receiver Operating Characteristics (ROC). Experimental results show that GNN-AD results in higher accuracy and larger Area Under the Curve (AUC) values compared to other supervised and unsupervised multi-stage end-to-end machine learning techniques.