基于社区检测的区块链分片算法OA

With the rapid growth in the scale of blockchain networks and transaction volume,traditional blockchain systems face scalability bottlenecks in terms of transaction throughput and confirmation latency.Sharding technology improves blockchain scalability by enabling parallel transaction processing.However,traditional random sharding methods based on account address hashing ignore transaction correlations among accounts,resulting in a high proportion of cross-shard transactions and degraded system performance.To reduce cross-shard transactions,we propose a temporal-sensitive sharding optimization algorithm based on Leiden community detection,namely Temporal-sensitive Sharding based on Leiden(TS-Leiden).The TS-Leiden algorithm employs community detection to mine the community structure of accounts and assigns accounts with strong transaction correlations to the same shard,thereby reducing the proportion of cross-shard transactions.In addition,a temporal-sensitive weight decay mechanism is introduced to reduce the long-term noise introduced by early historical transactions and enhance the representation of account transaction features.Experimental results show that the TS-Leiden algorithm effectively reduces the proportion of cross-shard transactions and transaction confirmation latency while im-proving system throughput,verifying the effectiveness of the proposed algorithm.