Performance Analysis of Dynamic Nodes Deployment Mechanism for Energy Consumption Optimization in Consortium Blockchain

As the consortium blockchain gradually emerges as the primary solution for enterprise-level trusted transactions, the issue of energy consumption during system operation has garnered increasing attention. The dynamic management capabilities supported by the consortium blockchain make it possible to dynamically adjust the scale of nodes participating in the blockchain operation based on the business scale. This paper first abstracts the key stages of consensus in the representative consortium blockchain, Hyperledger Fabric, into a queueing network model to simulate the interaction processes among different nodes in the consensus. Subsequently, based on the formal analysis of the model, a node sleeping and staggered wakeup mechanism for consortium blockchain is proposed, enabling the system to dynamically shut down and activate idle nodes, thereby improving the power consumption of the consortium blockchain while ensuring consensus latency. Experimental results show that with the introduction of node sleep and staggered wake-up mechanisms in the coalition chain, the average latency and throughput remain largely unaffected. However, this mechanism significantly reduces CPU and memory usage, as well as overall system energy consumption, under low load conditions.