Modeling and controlling spatiotemporal malware propagation in mobile Internet of Things

The mobility of devices in mobile Internet of Things (IoT) enables dynamic interactions, facilitating the spatiotemporal malware propagation. However, few studies have focused on accurately modeling and effectively controlling this form of malware propagation. To address this issue, we propose a theoretical framework that integrates patch-malware spreading dynamics with optimal patch allocation policy. First, we establish a novel temporal multilayer network comprising a central node, a patch dissemination layer, and a malware propagation layer. The hybrid patching process is implemented by the integration of the central node and the patch dissemination layer. In the malware propagation layer, the mobility of IoT devices is modeled as a diffusion process across multiple areas. Next, we design a dynamic spreading model to capture the evolution of malware propagation and analytically derive the invasion threshold. The threshold indicates that malware propagation is significantly influenced by both the patching process and the topological structure of mobile IoT. Furthermore, considering the central host's capacity and patch effectiveness, we develop an optimization algorithm to determine the optimal patch allocation policy under resource constraints. This algorithm significantly outperforms traditional centrality-based methods in malware mitigation. Finally, we analyze the impact of device mobility, the connectivity of the patch dissemination layer, the device distribution, the central node's capacity, and the patch effectiveness on malware propagation. Our study provides a theoretical foundation for predicting and controlling malware spreading in mobile IoT.