Trusted Microservice Orchestration for Secure Edge Computing in Industrial Cyber-Physical Systems

Md. Redowan Mahmud, Jiong Jin, Jonathan Kua, Mahbuba Afrin, Sajib Mistry, Aneesh Krishna

Abstract: The rapid penetration of Industrial Internet of Things (IIoT) and computing technologies has significantly transformed the landscape of Industrial Cyber-Physical Systems (ICPS). Recent disruptive paradigms, such as Cloud-Fog Automation, have advocated for the open deployment of ICPS at the cloud/fog layers. These paradigms aim to enhance system modularity and efficiency while leveraging the computing capabilities of cloud/fog technologies. The vast volume and diversity of data generated by ICPS can easily overwhelm traditional cloud solutions due to bandwidth and latency challenges. Edge-based analytics offers a promising solution by processing data closer to the source, yet deploying advanced analytics at the edge is challenging due to the computational, memory, and energy limitations of edge devices. Recently, microservices have emerged as an ideal candidate for providing lightweight and modular analytic services in ICPS. However, microservices introduce severe security vulnerabilities and operational complexities, particularly in heterogeneous and dynamic industrial settings. Existing microservice orchestration platforms are designed for stable cloud environments, hence they are unable to meet the unique security and adaptability requirements of industrial edge computing. To this end, we propose a novel vulnerability-aware microservice orchestration framework for secure edge computing in ICPS. We first conceptualize a trust model that detects behavioral anomalies in microservices, then perform extensive simulations using a state-of-the-art platform to validate the effectiveness of our trust model in enhancing cyber-physical security and system resilience in decentralized industrial edge computing environments.

External IDs:dblp:journals/network/MahmudJKAMK25