IEEE

Generative Pretraining Transformer Based False Data Injection Attack Detection Framework for DC Microgrid Under Uncertain Operating Condition

Abstract: Traditional False Data Injection Attack (FDIA) techniques typically rely on prediction-based models that forecast Renewable Energy Source (RES) generation. This reliance leads to increased ...
IEEE

Backdoor-Free Node Injection Attacks Against Graph Neural Networks

Abstract: Graph Neural Networks (GNNs) are vulnerable to backdoor attacks, where hidden triggers cause targeted misclassification. Existing attacks often require unrealistic control over the model or ...