UFETCH: A Unified Searchable Encryption Scheme and Its Saas-Native to Make DBMS Privacy-Preserving

Shen-Ming Chung,Ming-Der Shieh,Chia-Heng Tu,Tzi-Cker Chiueh,Chia-Chia Liu

doi:10.1109/access.2020.2994598

Abstract

As encrypted-search techniques such as Searchable Encryption (SE) were devised for homogeneous data type, i.e. textual or numerical, it is a nature presumption that multiple techniques have to be intertwined to make database management system (DBMS) privacy-preserving. In effect, such a presumption has led to popular designs such as CryptDB, putting efforts on heterogeneous integration. In this paper, an easier option is made available when a unified SE scheme named uFETCH is proposed to accommodate both. Namely, uFETCH is able to build unified SE indexes for both the types while enabling encrypted search even if the SE indexes built for texts and numbers are mingled. To demonstrate how uFETCH can bring up simpler designs, a security agent is exemplified to work with off-the-shelf DBMS while making it privacy-preserving. Since uFETCH transforms the problem of encrypted search into a simple problem of subsequence matching for SaaS-native, it requires only sub-linear search time w.r.t. the volume of indexed items and is secure in the widely-adopted 3-tier cloud structure to help cloud service providers ease regulation compliance with out-sourced repository.

Highlights

As General Data Protection Regulation (GDPR) [1] and similar laws are acting in more countries, people are much aware of their rights on privacy
It would make no sense to private data owners if the companies, working as a cloud service provider (CSP) in the era of cloud, further moves private data to yet another CSP for repository
We introduce a unified Searchable Encryption (SE) scheme called uFETCH that enables efficient encrypted-search across data types

Summary

INTRODUCTION

As General Data Protection Regulation (GDPR) [1] and similar laws are acting in more countries, people are much aware of their rights on privacy. Chung et al.: uFETCH: A Unified SE Scheme and Its Saas-Native to Make DBMS Privacy-Preserving. SaaS-native techniques, CryptDB is gaining popularity as it fits well with DaaS that forbids any DBMS retrofit Such an integration even inspired works such as MONOMI [33], L-EncDB [22] and Seabed [34]; all putting efforts on intertwining heterogeneous techniques that we propose to refrain from. With the merit of SaaS-native to make off-the-shelf DBMS privacy-preserving, uFETCH brings the following contributions to advocate cloud security and privacy. UFETCH builds SE trapdoors devised to tell nothing about the type of data it is looking for It features wildcard-based pattern search for encrypted texts while making range-based search possible for encrypted numbers. Cloud repository running as DaaS often provides DBMS not open for retrofit This fact cripples all encrypted-search techniques unless they are SaaS-native, i.e. able to work with DBMS as is. The operations and notations illustrate how the problem of encrypted search (using wildcards in particular) can be transformed into a problem of subsequence matching for SaaS-native

INDEX CONSTRUCTION

Findings

CONCLUSION