Predicting Propagation of Stealthy Worm for Unknown Source in Mobile Edge Computing

While improving network performance, Mobile Edge Computing (MEC) faces significant security threats due to User Equipment (UE) vulnerabilities and insecure communication protocols. This enables attackers to insert malicious nodes into the entire system and propagate worms such as the Mirai, enabling activities like DDoS attacks and cryptocurrency mining. To reduce security risks, it is necessary to identify worm sources and predict their propagation accurately. However, in practice, there exist a large number of stealthy worms (nodes that infect stealthy worms work without exhibiting any abnormal behavior), making it challenging to accurately analyze their source and predict their propagation. To address this problem, our paper introduces a prediction scheme designed to analyze the propagation of stealthy worms originating from unknown sources. First, a Bayes rule-based observation algorithm is designed to correct false-negative nodes and increase the accuracy of identifying the propagation sources. Then, we design a branch-and-bound algorithm to effectively reduce the traversal space and improve the efficiency of calculating the upper and lower bounds. Finally, a series of experiments in four real networks are conducted to show the accuracy and efficiency of our scheme.