Research on Hierarchical Sharding Based on Spectral Clustering Algorithm

Traditional blockchains face scalability issues such as low throughput, high transaction delay, and high storage redundancy. Introducing sharding technology can effectively improve the scalability of blockchain systems. However, current sharding techniques typically use random sharding, which may have an impact on the performance of the entire blockchain system. In response to the above issues, this article proposes a hierarchical sharding scheme based on spectral clustering and semi-sharding storage. This scheme adopts a master-slave hierarchical deployment of the blockchain, with the main chain responsible for integrating the state information of shard sub-chains, achieving consistency in the state of master-slave multi-chains. In the chain design, this article uses a spectral clustering algorithm to partition nodes with similar performance and geographical location into the same shard, thereby reducing transaction delay and ensuring performance balance among nodes within the shard. The simulation experiment results show that the proposed sharding scheme can significantly shorten transaction delay and effectively improve the throughput of the blockchain system.