Blockchain-Enabled Task Offloading With Energy Harvesting in Multi-UAV-Assisted IoT Networks: A Multi-Agent DRL Approach

【Author】 Seid, Abegaz Mohammed; Lu, Jianfeng; Abishu, Hayla Nahom; Ayall, Tewodros Alemu

【Source】IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS

【影响因子】13.081

【Abstract】Unmanned Aerial Vehicle (UAV) is a promising technology that can serve as aerial base stations to assist Internet of Things (IoT) networks, solving various problems such as extending network coverage, enhancing network performance, transferring energy to IoT devices (IoTDs), and perform computationally-intensive tasks of IoTDs. Heterogeneous IoTDs connected to IoT networks have limited processing capability, so they cannot perform resource-intensive activities for extended periods. Additionally, IoT network is vulnerable to security threats and natural calamities, limiting the execution of real-time applications. Although there have been many attempts to solve resource scarcity through computational offloading with Energy Harvesting (EH), the emergency and vulnerability issues have still been under-explored so far. This paper proposes a blockchain and multi-agent deep reinforcement learning (MADRL) integrated framework for computation offloading with EH in a multi-UAV-assisted IoT network, where IoTDs obtain computing and energy resources from UAVs. We first formulate the optimization problem as the joint optimization problem of computation offloading and EH problems while considering the optimal resource price. And then, we model the optimization problem as a Stackelberg game to investigate the interaction between IoTDs and UAVs by allowing them to continuously adjust their resource demands and pricing strategies. In particular, the formulated problem can be addressed indirectly by a stochastic game model to minimize computation costs for IoTDs while maximizing the utility of UAVs. The MADRL algorithm solves the defined problem due to its dynamic and large-dimensional properties. Finally, extensive simulation results demonstrate the superiority of our proposed framework compared to the state-of-the-art.

【Keywords】Bkckchain; computation offloading; energy harvesting; Internet of Things; MADRL; UAV

【发表时间】2022 DEC

【收录时间】2023-01-14

【文献类型】实验仿真

【主题类别】

区块链应用-实体经济-无人机领域