Building Advanced and Versatile Applications over Encrypted Data

Abstract

Moving ever-growing data to the commercial public cloud for storage, processing, and management has become increasingly popular, due to the prominent benefits of cloud computing like scalability, economical cost, and ubiquitous network access. While being highly beneficial, deploying data-driven applications on the cloud which is typically operated by external third-parties also raises severe privacy concerns. Meanwhile, the increasingly strict legal regulations also pose the demand for the promise of privacy assurance in cloud-based applications. Ideally the cloud should serve without seeing the data in clear, yet simply applying standard encryption would prevent effective data utilization.

This dissertation presents algorithmic design for building advanced and versatile applications over encrypted data in cloud computing, accommodating data types ranging from simple numerical sensory measurement to the complex case of images and videos. Firstly, we propose the design of a cloud-based system for privacy-preserving truth discovery applications, which aims to mine in a privacy-friendly manner truthful information from unreliable multi-source sensory measurement in cloud-enabled mobile crowdsensing systems. Secondly, we propose the design of a cloud-based system for privacy-preserving image denoising applications, which enables the cloud hosting encrypted image databases to provide secure query-based image denoising services for the generation of high quality image content. Thirdly, we propose the design of a cloud-based system for secure video deduplication applications with adaptive video delivery, which arms the cloud with the crucial deduplication functionality to eliminate the storage and bandwidth redundancy due to hosting encrypted videos from different entities, as well as with the support of adaptive video dissemination for heterogeneous networks and devices. The presented research would greatly advance the adoption of cloud computing in our daily life to safely serve versatile data service demands.

Date of Award	24 Apr 2019
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Cong WANG (Supervisor)