Blockchain-Secured Online Edge Collaboration in IoT: Integrating Convex Optimization and Learning Approach

• Published in: IEEE internet of things journal Vol. 12; no. 21; pp. 45376 - 45392
• Main Authors: Xu, Yueqiang, Liu, Zhi, Li, Jiaxin, Jiang, Jing, Zhang, Heli, Liu, Zhibin, Lin, Fuhong
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Blockchain; Blockchains; Collaboration; Convex analysis; Convexity; Cybersecurity; Deep learning; deep reinforcement learning (DRL); Dynamic scheduling; edge collaboration; Edge computing; Energy consumption; Internet of Things; Internet of Things (IoT); Optimization; Queuing theory; Resource allocation; Resource management; Security; Servers; Stability; Task scheduling; Utility theory
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2025.3600299

Edge collaboration has emerged as a promising paradigm for Internet of Things (IoT) applications. However, achieving efficient cooperation among these server nodes still faces several critical challenges, including: 1) secure node interaction; 2) online task scheduling; and 3) heterogeneous resource management. Unfortunately, most existing solutions address these issues in isolation, lacking an integrated framework that jointly considers security, task scheduling, and resource management. To address these limitations, this article proposes a blockchain-based online collaboration framework for IoT, where blockchain serves as a trusted top-layer management platform to ensure secure information sharing and resource management. In the proposed framework, we introduce two dynamic queues to effectively manage randomly arriving tasks and develop an online collaboration mechanism tailored for heterogeneous edge servers. Furthermore, we formulate a long-term system utility maximization problem by jointly optimizing collaboration strategies, resource allocation, and block producer selection, subject to queue stability and security constraints. Due to the coupling among decision variables and across time slots, solving the optimization problem directly is challenging. Therefore, we design a novel Lyapunov-based algorithm that integrates convex optimization theory with deep reinforcement learning (DRL), significantly improving the solving efficiency. Extensive simulations demonstrate that the proposed method and algorithm outperform conventional baseline methods and pure DRL-based approaches in terms of system utility, stability, and security performance, making it a promising solution for secure and efficient edge collaboration in dynamic IoT environments.