Privacy-enhancing Cloud Computing Solution for Big Data

Privacy-enhancing Cloud Computing Solution for Big Data

Stať ve sborníku v databázi WoS či Scopus

A variety of users cyber data are daily collected by enterprises and governments. These data are processed in order to improve users lifestyle, e.g. improving healthcare and optimizing traffic. Cloud computing is often the only possible strategy which allows processing these big amount of information due to the associated costs. However, data owners remain reluctant to outsource their data to a cloud which may read, use or even sell these data leading on user privacy leakage. These threats have been already observed by EU organization and reflected in many regulations and strategies. In this paper, we present data splitting techniques as one of the most promising technology for privacy-preserving computation in a cloud. At first, we propose a new protocol for secure scalar product resistant against honest-but-curious colluding cloud. Then, we compared our new proposal with two most efficient state-of-the-art schemes. These schemes are not secure in a colluding security model. While using higher security level of our proposal, we achieved comparable performance. The protocols were implemented in real environment based on Amazon Web Services and, then, an FPGA processor cards implementation is considered in order to speed up the computations.

A variety of users cyber data are daily collected by enterprises and governments. These data are processed in order to improve users lifestyle, e.g. improving healthcare and optimizing traffic. Cloud computing is often the only possible strategy which allows processing these big amount of information due to the associated costs. However, data owners remain reluctant to outsource their data to a cloud which may read, use or even sell these data leading on user privacy leakage. These threats have been already observed by EU organization and reflected in many regulations and strategies. In this paper, we present data splitting techniques as one of the most promising technology for privacy-preserving computation in a cloud. At first, we propose a new protocol for secure scalar product resistant against honest-but-curious colluding cloud. Then, we compared our new proposal with two most efficient state-of-the-art schemes. These schemes are not secure in a colluding security model. While using higher security level of our proposal, we achieved comparable performance. The protocols were implemented in real environment based on Amazon Web Services and, then, an FPGA processor cards implementation is considered in order to speed up the computations.

Cloud Computing; Data Splitting; Privacy; Big Data; FPGA

Cloud Computing; Data Splitting; Privacy; Big Data; FPGA

SMÉKAL, D.; RICCI, S.; DZURENDA, P.; MARTINÁSEK, Z.

2020

28.10.2019

IEEE Computer Society

Dublin, Ireland

9781728157634

2019 1th International Congress on Ultra Modern Telecommunications and Control Systems (ICUMT)

2157-023X

International Congress on Ultra Modern Telecommunications and Workshops

Spojené státy americké

1

6

6

https://ieeexplore.ieee.org/abstract/document/8970982