.png){kind=icon}
.png){kind=icon}
Return
Dynamic sharding model and performance optimization method for consortium blockchain
.png){kind=icon}
PRE
AI
DOI：10.1007/s11227-024-06870-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Consortium blockchain has been widely used in finance, e-government due to the characteristics of controllability, supervision, and operability. However, traditional consortium blockchains have bottlenecks in throughput. To solve the bottlenecks, the paper proposes a Dynamic Sharding Model and Performance Optimization Method for Consortium Blockchain (DSPO-CB), which offers a new shard architecture and dynamically optimizes the architecture through the Deep Q-Network (DQN). Firstly, the model reduces redundancy and improves space utilization by classifying the nodes ensuring security. Secondly, the model proposes the shard structure through a dynamic clustering method based on the node status to reduce the proportion of cross-shard transactions. Finally, the DQN is used to dynamically optimize the sharding and consensus architecture. Experiments show that DSPO-CB improves the throughput by 33% and saves up to 78% storage space compared with the existing consortium blockchain.
Keywords:
Consortium blockchain
Dynamic sharding
Performance optimization
Reinforcement learning
Journal
IF:
2.7
Papers:
990
Citations:
1.0W