Energy-efficient blockchain-IIoT with mobile edge computing: optimizing resource allocation and multi-hop offloading

Abstract

Blockchain and Mobile Edge Computing (MEC), both of which are edge computing paradigms characterized by strong security attributes, have become a research focus in the Industrial Internet of Things (IIoT) domain. However, current studies generally overlook the persistent energy consumption burden caused by the computational and mobile communication components required for these architectures, which further intensifies the energy supply constraints faced by IIoT devices. To this end, this article proposes an energy-efficient blockchain-enabled IIoT integrated with MEC to optimize resource allocation and multi-hop task offloading. Firstly, an innovative Modular Industrial Internet of Things Device (MIITD) is designed to reduce sustaining energy consumption by dynamically disabling non-essential computational and communication modules during multi-hop offloading. Secondly, a blockchain-MEC architecture is structured to incorporate multiple system-level variables, including multi-aggregation MIITDs deployment, offloading strategies, block size, and wireless link weights. Thirdly, a joint optimization framework is developed to prolong system lifetime, minimize packet loss, and manage computational overhead. Finally, a Multi-Agent Hybrid Trust Region Policy Optimization (MAHTRPO) algorithm is proposed to optimally solve the formulated problem. Experimental and simulation results validate that the proposed scheme substantially enhances system endurance and ensures low data packet loss rates in MIITD scaling scenarios.

Keywords

• Industrial Internet of Things (IIoT)
• Blockchain
• Mobile Edge Computing (MEC)
• Multi-hop computation offloading
• Multi-agent deep reinforcement learning