Verifiable data search atop blockchain

Abstract

As an emerging decentralized technology, blockchain has become a compelling paradigm for trusted data storage owing to the underlying techniques of hashing chain and consensus schemes. In addition to the on-chain data itself, blockchain can also be utilized to store key data outsourced from data owners via smart contract. It can be seen as an important complement and enhancement to existing cloud storage. However, untrusted clouds necessitate the verifiable data search atop the blockchain. Apart from blockchain data, raw data or encrypted data can also be outsourced to the cloud. Therefore, in this thesis, we investigate three types of data, i.e., native blockchain data, outsourced raw data and outsourced encrypted data, in the scenarios containing both clouds and blockchain. For the native blockchain data, we employ clouds to provide efficient query services on the underlying data and design a Verifiable Query Layer (VQL) to make the query verifiable. In terms of outsourced data, we let clouds store the data and host query services over it. The blockchain will store some metadata via the smart contract and facilitate the query verification. For the outsourced raw data, we focus a complicated data structure, i.e., graph data, and enable privacy-preserving verifiable query by designing a novel authenticated data structure (ADS) named PAGB. To handle outsourced encrypted data, we propose a novel verifiable searchable symmetric encryption (SSE) scheme called Slicer to support range search on numerical data. The effectiveness and practicality of our designs are demonstrated by theoretical analysis and extensive evaluations respectively.