区块链动态分片自适应模型

Dynamic sharding adaptive model for blockchain

分片机制是解决区块链吞吐量低和通信复杂度高等性能瓶颈问题的一种途径。然而在现有的分片系统中,分片之间的交易可以并行处理,这样会不可避免地出现频繁的跨分片交易,相比于片内交易,跨分片交易会增加确认时间。当系统中跨分片交易增多时会导致分片系统效率降低。提出了一种动态分片自适应模型,用于提高区块链平台的吞吐率和效率。将通过交易数据相关性动态地将数据拆分到同一个分片,同时将多个高度相关的分片自适应地合并到同一个分片当中。为了验证区块链动态分片自适应模型,使用超级账本对区块链动态分片自适应模型进行实验。实验结果表明所提出的模型可以提升区块链分片系统的可扩展性,同时保证安全性以及一致性问题。

Sharding mechanism is a way to address performance bottlenecks in blockchain such as low throughput and high communication complexity.However,in existing sharding systems,transactions across shards can be processed in parallel,inevitably leading to frequent cross-shard transactions which increase confirmation times compared to intra-shard transactions.As the number of cross-shard transactions increases,the efficiency of the sharding system decreases.This paper proposed a dynamic sharding adaptive model to enhance the throughput and efficiency of blockchain platforms.By dynamically splitting data into the same shard based on transaction data correlation,multiple highly correlated shards were adaptively merged into the same shard.To validate the dynamic sharding adaptive model for blockchain,it conducted experiments using Hyperledger.Experimental results demonstrate that the proposed model can enhance the scalability of blockchain sharding systems while ensuring security and consistency.