A Fine-Grained Verification Method for Blockchain Data Based on Merkle Path Sharding

Liang Wen, Zhiqiong Wang, Tingyu Cui, Caiyun Shi, Baoting Li, Zhongming Yao

Abstract: Blockchain has been widely applied these years. As its volume of data is greatly increasing, it is necessary to shard the data to utilize hardward storage efficiently. After sharding, however, the validated data are stored in different nodes and read cross-shardingly, which is inefficient because of extra communication time. To solve this problem, the authors use the Mp-tree (Merkle Path Tree) structure to replace the Merkle tree in the traditional blockchain, and then shard the blocks according to the Mp-tree architecture. The Mp-tree can validate the accuracy of data to avoid cross-sharding so as to improve the efficiency of validation. It has some advantages: first, it can reduce the volume of storage space by fine-grained slicing; second, it can improve the validation efficiency of blockchain slicing by independent validation. This study demonstrates that our design can optimize the generation efficiency in existing sharding systems, and also exploiting its independent verification feature to improve verification efficiency.