Research On Verifiable Query Optimization For Blockchain

In recent years,the rapid development of blockchain technology has led to an increasing demand for querying on-chain data in various scenarios.In most public blockchains,the majority of users participate in the network as light nodes due to limited storage and computational resources.However,light nodes can not process query individually,for they only hold block headers,so they must query full nodes for results when there is a query demand.As the nodes in a blockchain network are mutually untrusted,the blockchain needs to provide verifiable query services for light nodes to ensure the integrity of query results.Existing verifiable query solutions often require additional computation or transmission of large amounts of verification objects,which severely impacts their practicality.This study aims to solve the performance problem of verifiable queries on blockchain and proposes two efficient and lightweight schemes for verifiable boolean range queries.The detailed research contents of this paper are as follows:1.A merkle tree-based verifiable query scheme is proposed to enhance query performance and reduce query costs for light nodes.To solve the performance problem,the scheme divides the query process into two stages:inter-block filtering and intrablock querying.For the intra-block querying stage,we design a simplified Merkle-B tree that speeds up query efficiency while reducing the size of verification objects generated for each block.For inter-block filtering,we ingeniously combine the Merkle tree with Bloom filters and bitmaps to propose two efficient authenticated structures that can filter out blocks without results in batch and accelerate query processing.We also propose an optimization based on fixed windows to further improve query performance.Experimental results show that the merkle tree-based verifiable query scheme improves query performance by 98%at most and reduces the size of verification objects by 58%compared to existing solutions.2.To address the limited network bandwidth problem in scenarios where mobile devices act as light nodes,we propose a communication-friendly verifiable query scheme based on vector commitment technology,aiming at further reducing the size of transmitted verification objects.This scheme designs a new verifiable structure called Verkle-B tree,which is based on vector commitment and B+tree,to reduce generated proofs by increasing fan-out,and uses vector commitment to optimize inter-block authenticated structures.To mitigate the performance impact caused by vector commitment,we also propose an aggregation method for multi-proofs and an asynchronous verification scheme.Experimental results show that the vector commitment-based verifiable query scheme further reduces the transmitted verification size by 17%and improves query response speed of mobile device by 14%compared to the merkle treebased scheme.