摘要
近年来,IIoT(工业物联网)的发展十分迅速,其应用场景也极为丰富,如传感器、机器人、自动化设备等[1]。工业物联网在实际应用中会产生大量的数据,数据的隐私性及安全性是十分重要的。目前,区块链技术被广泛地认为是实现数据安全高效存储/处理/共享的一种有前途的解决方案,但区块链在扩展性、去中心化程度、安全性及延迟上四个方面存在着相互制约。利用深度强化学习可以更好地在上述四个方面实现较好的平衡,使得区块链能够满足工业物联网的吞吐量的同时有较好的其他性能。
In recent years,IIoT(Industrial Internet of Things)has developed rapidly,and its application scenarios are extremely rich,such as sen⁃sors,robots,automation equipment,etc.At the same time,the industrial Internet of things will produce a lot of data in practical application,the privacy and security of data is very important.At present,blockchain technology is widely considered as a promising solution to realize data safe and efficient storage/processing/sharing.However,there are four constraints on the blockchain in terms of scalability,decentral⁃ization,security and latency.The use of deep reinforcement learning can find a better balance in the four aspects,so that the blockchain can meet the throughput of the industrial Internet of things and have better other performance.
作者
王煜翔
WANG Yuxiang(College of Computer Science,Sichuan University,Chengdu 610065)
出处
《现代计算机》
2021年第9期19-23,共5页
Modern Computer
关键词
IIoT
区块链
强化学习
性能
IIoT
Blockchain
Reinforcement-Learning
Performance