面向工业互联网的区块链分层分片研究

Research on Layering and Sharding of Blockchain for Industrial Internet

区块链的不可篡改、去中心化等特点能够有效解决工业互联网中日益突出的安全和隐私问题,然而当前主流区块链平台的吞吐量远不能满足工业互联网海量数据快速上链的需求,并且传统区块链采用的高冗余存储机制也无法适用于工业互联网场景。建立分层分片区块链架构,将区块数据分层存储在多个分布式云服务器和边缘服务器中,以应对工业互联网不断增长的数据量。依据边缘服务器之间的拓扑结构,设计一种基于复杂网络社团划分算法的改进区块链网络分片算法,在提升区块链网络吞吐量的同时缩短分片时间。将区块广播过程形式化为生成树论证了区块广播时间对吞吐量的影响,在此基础上提出一种基于生成树的片内主节点选取算法,进一步提升区块链网络的吞吐量。实验结果表明,与经典复杂网络社团划分算法相比,改进的区块链网络分片算法在对大规模网络进行分片时能够在不牺牲分片质量的前提下缩短约36%的分片时间,同时减少了各分片内区块的广播时间。

The blockchain technology can effectively solve the increasingly prominent security and privacy issues in the industrial Internet because its non-tampered and decentralized characteristics.However,the throughput of the current mainstream blockchain platform is far from satisfying the requirement for rapid on-chaining of massive data in industrial Internet,and the high redundancy storage mechanism adopted by the traditional blockchain can not be applied to industrial scenarios.A layering and sharding blockchain architecture is established.The block data is layered and stored in multiple distributed cloud servers and edge servers to cope with the increasing amount of data in the industrial Internet.Based on the topology between the edge servers,an improved blockchain network sharding algorithm based on complex network community division algorithm is designed to improve the throughput of the blockchain network and shorten the sharding time.The block broadcast process is formalized as a spanning tree,and the influence of block broadcast time on throughput is demonstrated.Based on this,a leader-node selection algorithm in the shard based on spanning tree is proposed to further improve the throughput of the blockchain network.The experimental results show that compared with the classical complex network community division algorithm,the improved blockchain network sharding algorithm can shorten the sharding time by approximately 36%without sacrificing the sharding quality and simultaneously reduce the broadcast time of blocks in each shard.