Building a Secure and Practical Decentralized Storage Auditing Framework

構建安全實用的去中心化存儲審計框架

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date3 May 2022

Abstract

How to audit outsourced data in centralized storage like cloud is well-studied in the past decade, but it is largely under-explored for the rising decentralized storage (DS), which is widely expected to bode well for a future market worth billion dollars. In contrast to centralized storage, DS fulfills the emerging user demands of storage decentralization with a fully incentivized mechanism. To realize these demands, the blockchain technology plays an essential and indispensable role in recording and verifying audit trails in forms of proof of storage. In addition, the blockchain is also used to handle fair payments with necessary dispute resolution. However, the inherent limitations of blockchain technology also inevitably lead to numerous challenges that conventional proof-of-storage solutions cannot resolve at once.

In light of the rapid development of DS, this dissertation presents a comprehensive study on the current state-of-the-art DS storage auditing designs by the end of 2021. We identify key security problems as well as practicality constraints, and propose feasible designs and implementation to enable secure, convenient, and dynamic storage auditing for DS. Concretely, we first present a basic storage auditing solution that addresses the on-chain privacy concerns stemmed from the transparency of blockchain. This basic solution demonstrates a secure approach to form up an ecosystem of decentralized storage with a compatible incentive mechanism. Secondly, in face of the constraints of high audit cost and overwhelming computational overhead required for decentralized storage auditing, we propose a middleware-empowered solution to facilitate off-chain procedures with the assistance of trusted execution environment. Lastly, we extend to a more useful setting of dynamic data storage and we propose an efficient storage auditing design in support of mutable storage for DS. We hope the presented research would advance the understanding of decentralized storage auditing, and encourage new techniques and designs for a more promising decentralized storage marketplace.