CD-PBFT: Incentive-based Efficient Blockchain Consensus Mechanism for Web 3.0.

Web 3.0 is the promising technology which enables users to read, write, and own user-generated data. It is attracting attentions from academia and industries to based on Web 3.0 construct infrastructures to provide services for participants. Blockchain is a key technology to implement Web 3.0, which can record data and prevent data mutations in a secure and decentralized way. Since Web 3.0 is an integration of multiple technologies, blockchain is required to frequently interact with them, which challenges the throughput of consensus. An efficient consensus algorithm plays an important role to support Web 3.0 infrastructures and services. However, current researches on consensus algorithms require large-scale nodes to participate and switch view frequently, which causes high communication overheads and long consensus delays. To solve this problem, in this paper, we propose an incentive-based blockchain consensus mechanism for Web 3.0, called CD-PBFT, to improve efficiency and record consensus quality to regulate behaviors of nodes. Since nodes with higher incentives and qualities are selected as the primary nodes frequently to reduce view switching, a cooling mechanism is introduced to prevent these nodes dominating the consensus process. Moreover, a K-medoids-based clustering mechanism is elaborated to form a multi-cluster network to reduce the amount of nodes required for consensus. Simulation results prove that the proposed consensus mechanism can prevent centralization on the primary nodes, and reduce communication overheads and consensus delays.