Proof of Similarity: Searchability-enhancing Blockchain Consensus for Industrial Digital Twins

Recently, industrial digital twins have been widely adopted to create a digital replica of the industrial sites for real-time simulation, intelligent decision-making, and automatic feedback control. However, the bi-directional information flows between the physical twin and digital twin are facing unique challenges in guaranteeing data authenticity and improving data searchability. Existing studies propose employing blockchain-empowered industrial digital twins to authenticate the information flow. They also consider improving data security and privacy and deploying blockchain-empowered industrial digital twins on emerging 6G and edge networks. However, they neglect the provision of high-efficiency data query services. This work is the first to consider the searchability issue in blockchain-empowered industrial digital twins. We propose a general system architecture of blockchain-empowered industrial digital twins applicable for various applications. The key novelty of this work lies in the proof-of-similarity blockchain consensus mechanism that significantly improves the search efficiency of industrial digital twins data. More specifically, proof of similarity guarantees that the generated block contains semantically similar data and a block summary. Then, the blockchain queries will be compared with the block summaries before entering the transactions in the blocks. In this way, up to $\mathbf{9 2. 0 \%}$ of the blocks can be skipped during blockchain query, resulting in an 87.4% reduction of the average query completion time.