Iquery: A Trustworthy and Scalable Blockchain Analytics Platform
Lingling Lu,Zhenyu Wen,Ye Yuan,Binru Dai,Peng Qian,Changting Lin,Qinming He,Zhenguang Liu,Jianhai Chen,Rajiv Ranjan
DOI: https://doi.org/10.1109/tdsc.2022.3228908
2023-01-01
IEEE Transactions on Dependable and Secure Computing
Abstract:Blockchain, a distributed and shared ledger, provides a credible and transparent solution to increase application auditability by querying the immutable records written in the ledger. Unfortunately, existing query APIs offered by the blockchain are inflexible and unscalable. Some studies propose off-chain solutions to provide more flexible and scalable query services. However, the query service providers (SPs) may deliver fake results without executing the real computation tasks and collude to cheat users. In this article, we propose a novel intelligent blockchain analytics platform termed iQuery , in which we design a game theory based smart contract to ensure the trustworthiness of the query results at a reasonable monetary cost. Furthermore, the contract introduces the second opinion game that employs a randomized SP selection approach coupled with non-ordered asynchronous querying primitive to prevent collusion. We achieve a fixed price equilibrium, destroy the economic foundation of collusion, and can incentivize all rational SPs to act diligently with proper financial rewards. In particular, iQuery can flexibly support semantic and analytical queries for generic consortium or public blockchains, achieving query scalability to massive blockchain data. Extensive experimental evaluations show that iQuery is significantly faster than state-of-the-art systems. Specifically, in terms of the conditional, analytical, and multi-origin query semantics, iQuery is 2 ×, 7 ×, and 1.5 × faster than advanced blockchain and blockchain databases. Meanwhile, to guarantee 100% trustworthiness, only two copies of query results need to be verified in iQuery , while iQuery 's latency is $2 \sim 134$ × smaller than the state-of-the-art systems.