A Trusted Storage System for Digital Object in the Human-Cyber-Physical Environment
Xiang Jing,Yueyang Hu,Chaoran Luo,Xingchun Diao,Gang Huang,Haiou Jiang
DOI: https://doi.org/10.1007/978-981-19-8043-5_3
2022-01-01
Abstract:The continuous development of the information technology has driven the deep integration of various networks such as the Internet, Internet of Things, and mobile networks, promoting the fusion of human society, information space, and the physical world and the development of the human-cyber-physical (HCP) environment. In the HCP environment, applications usually need to use heterogeneous resources belonging to multiple systems and participants. The Digital Object Architecture (DOA), a data-centric software architecture proposed by Dr. Robert Kahn and aims to enable interoperability across systems, is useful in the HCP environment. HCP resources can be uniformly abstracted and modeled as digital objects (DOs), and all the operations and interactions of HCP resources can be transformed into the DO operations and interactions. There are two challenges of using DOA in the HCP environment, the first is the trust DO governance because of the "digital" nature of DO, making itself easily copied; the second is the dynamic characteristics of HCP resources, making it necessary to store the history state of resource DOs. In this paper, a trusted storage system for resource DOs in the HCP environment is proposed. By using blockchain technology, the change records of DOs can be stored in blockchain nodes, maintained by committee constituted by authoritative participants in the HCP environment. Two algorithms are proposed in the block generation and consensus stage to improve the whole performance of the system. In the block generation stage, the Periodic-Adjusted Merkle (PA-Merkle) tree is proposed and reduces time consumption by 27.8%. In the consensus stage, the Tree Spread PBFT (TS-PBFT) algorithm is proposed and reduces the time consumption by more than 20% under 100 Mbps bandwidth. The overall performance of the system improves by 24.2% in the case of four nodes with 100 Mbps bandwidth.