Research on Optimal Scheduling Algorithm Based on Dynamic Load Balancing of Alliance Blockchain Network

Hyperledger Fabric (Fabric), a fully operational consortium blockchain platform that supports smart contracts, has seen widespread adoption in various commercial projects. As Fabric’s popularity and real-world application grow, meticulous planning and deployment of its network configuration become crucial to meeting the throughput and latency requirements developers demand. In a multi-channel setting, while Fabric safeguards interactions among nodes sharing common objectives but operating at different trust levels, it employs traditional fault-tolerant consensus protocols relying on node identities to avoid costly mining operations. However, this approach does not entirely eliminate the risk posed by malicious nodes, threatening the system’s security and efficiency. To address these challenges, we propose a Transaction Batch Scheduling Algorithm (TBSA), specifically tailored for multi-channel architectures, aiming to maximize system utility and mitigate transaction backlog by identifying the optimal batch composition, particularly when dealing with future transactions destined for uncertain locations. Our TBSA algorithm was implemented on a cloud server based on the stable version of Fabric v2.0. Its efficacy was corroborated through experimental validation, which determined the ideal transaction batch sizes at various intervals and assessed the consequent system utility and delay distribution across different numbers of channels. These findings affirm the potential of the TBSA algorithm in enhancing system performance.