期刊
IEEE WIRELESS COMMUNICATIONS
卷 28, 期 2, 页码 172-180出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/MWC.001.2000102
关键词
Blockchain; Computer architecture; Resource management; Wireless communication; Edge computing; Wireless sensor networks; Smart contracts
类别
资金
- National Natural Science Foundation of China [U1633115, 61671029]
- Science and Technology Foundation of the Beijing Municipal Commission of Education [KM201810005027]
- Natural Science Foundation of Beijing Municipality [L192002]
- Open Foundation of the Key Laboratory of Airport Group Intelligent Operation [KLAGI020181004]
This article introduces a new architecture for resource sharing and transactions in fog computing networks, called B-ReST, enabled by blockchain technology. Key technologies and advantages of B-ReST are discussed, and simulation results demonstrate the benefits of B-ReST in resource sharing and transactions through solving the RPM problem.
Driven by the extensively emerging applications requiring big data processing, a series of heterogeneous network architectures have been proposed to meet user experience requirements. Among them, the concept of fog computing facilitates the effective integration and utilization of ubiquitous computing resources. In fog computing scenarios, willingness and service billing issues become significant to computing resource sharing and transactions. In this article, the recently developed blockchain technology characterized by successfully enabling consensus in an untrustworthy environment is introduced. Based on the blockchain technology, we propose a new architecture for resource sharing and transactions in fog computing networks, named Blockchain-Enabled Resource Sharing and Transactions in Fog Computing (B-ReST). The physical architecture, functional architecture, and workflow in B-ReST are defined. We also discuss the key technologies in B-ReST such as the smart contracts, the consensus mechanism and the requester and provider matching (RPM). The wireless characteristics of fog computing and blockchain technology are closely combined to make full and efficient use of ubiquitous computing resources. To prove the feasibility of the proposed architecture, the RPM problem is solved by a deep reinforcement learning (DRL) based method. Simulation results show the advantages of B-ReST to realize resource sharing and transactions, and the performance of B-ReST based on the DRL method has been enhanced. Challenges and future research directions are summarized as well.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据