Privacy Issues in Cloud computing Zeeshan Ali Shah System - - PowerPoint PPT Presentation

▶
privacy issues in cloud computing
SMART_READER_LITE
LIVE PREVIEW

Privacy Issues in Cloud computing Zeeshan Ali Shah System - - PowerPoint PPT Presentation

Aug 24, 2022 167 likes •356 views

Privacy Issues in Cloud computing Zeeshan Ali Shah System administrator PhD researcher KTH PDC Center for High Performance Computing W-Sunday, October 28, 12 Agenda Zeeshan Ali Shah Introduc5on to Cloud compu5ng

slide-1
SLIDE 1

Privacy Issues in Cloud computing

Zeeshan Ali Shah System administrator PhD researcher KTH PDC Center for High Performance Computing

W-Sunday, October 28, 12

slide-2
SLIDE 2

Agenda

  • Zeeshan ¡Ali ¡Shah

– Introduc5on ¡to ¡Cloud ¡compu5ng – Privacy ¡through ¡distribu5ng ¡data ¡

  • Ali ¡Gholami

– Privacy ¡in ¡Cloud ¡Compu5ng ¡Through ¡Iden5ty ¡ Management

  • QA

2

W-Sunday, October 28, 12

slide-3
SLIDE 3

3

FEATURES MODES LOCALITY BENEFITS COMPARES ¡TO STAKEHOLDERS

Reliability

TYPES

Elasticity Virtualisation … IaaS PaaS SaaS Public Private Hybrid … Local Remote Distributed Cost ¡Reduction Ease ¡of ¡use … Internet ¡of Services Grid Service-­‑oriented Architecture Resellers Providers Adopters Users …

Cloud ¡ Systems

"Advances in Clouds" , Expert Group Report (2012), Editors Lutz Schubert, Keith Jeffery

Big picture

W-Sunday, October 28, 12

slide-4
SLIDE 4

Private / Internal Cloud

Public Cloud

On Premises / Internal Off Premises / Third Party

Public / External

Hybrid

Deployment models

W-Sunday, October 28, 12

slide-5
SLIDE 5

5

Cloud Infrastructure (e.g. Physical/Virtual Hardware) Cloud Storage (e.g. Database) Cloud Platform (e.g. App Server) Cloud Services (e.g. Web Services) Cloud Application (e.g. SaaS) Cloud Clients (e.g. Browsers, Mobiles, Devices)

Armbrust, M., Fox, M., Griffith, R., et al. (2009) “Above the Clouds: A Berkeley View of Cloud Computing”, In: University of California at Berkeley Technical Report no. UCB/EECS-2009-28, pp. 6-7, February 10, 2009

User View

W-Sunday, October 28, 12

slide-6
SLIDE 6

6

Cloud Infrastructure (e.g. Physical/Virtual Hardware) Cloud Storage (e.g. Database) Cloud Platform (e.g. App Server) Cloud Services (e.g. Web Services) Cloud Application (e.g. SaaS) Cloud Clients (e.g. Browsers, Mobiles, Devices)

Lock-in

User View

W-Sunday, October 28, 12

slide-7
SLIDE 7

User Want bare metal resources Want high level tools to create my own services Want ready to go services

SaaS user PaaS = SaaS provider = IaaS user IaaS

Web applications Utility computing

How to chose: The Cloud Mindmap

W-Sunday, October 28, 12

slide-8
SLIDE 8

Privacy issues, Why ?

  • Out sourcing of compute
  • Out sourcing of Data
  • Multi Tenant
  • Assurance for Cloud resource provider
  • Cross border regulations
  • ..

W-Sunday, October 28, 12

slide-9
SLIDE 9

Some ways ..

  • Encryption
  • Anonymization
  • see next

W-Sunday, October 28, 12

slide-10
SLIDE 10

Distribute the data ..

  • Keeping Data Private while

Computing in the Cloud 2012 IEEE

Fifth International Conference on Cloud Computing (see in supplement reading)

W-Sunday, October 28, 12

slide-11
SLIDE 11

The paper describes

  • Preserving data privacy while solving comput-

intensive problems on un-trusted machines..

  • With help of Data Distribution in chunks
  • It outperforms existing crypto-based privacy

such as homomorphic encryption ---See results

W-Sunday, October 28, 12

slide-12
SLIDE 12

using tile assembly

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 (a) 1 1 1 1 1 1 1 1 1 1 (b) 1 1 1 1 1 1 1 1 1 1 1 1 (c) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 (d) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ` 1 ` 2 ` 3 ` 4 ` 5 ` 7 ` 6 ` 8 (e) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ` 1 ` 2 ` 3 ` 4 ` 5 ` 7 ` 6 ` 8 (f)

W-Sunday, October 28, 12

slide-13
SLIDE 13

Overview of sTile Algorithm

Initialization (Client) Initialization (Client) Replication (All Nodes) Replication (All Nodes) Recruitment (All Nodes) Discovery (All Nodes) Discovery (All Nodes) Solution (Client)

Figure 4. Overview of sTile algorithms.

W-Sunday, October 28, 12

slide-14
SLIDE 14

Privacy Preservation

  • Each tile being exposed only to a few

intermediate

  • A special tile encodes the solution, but has no

knowledge of the input

W-Sunday, October 28, 12

slide-15
SLIDE 15

Performance

Prob. # of Nodes Delay Execution Time Mahjong A 11 Private Cluster 20.1 sec. HPCC 19.3 sec. PlanetLab 18.5 sec. B 11 Private Cluster 41.6 min. HPCC 41.2 min. PlanetLab 43.9 min. Simjong D 1,000,000 0ms 65 min. 10ms 57 min. 100ms 64 min. 500ms 60 min. Gaussian 68 min. Distance-based 59 min.

Figure 7. The effect of network delay on system execution time.

W-Sunday, October 28, 12

slide-16
SLIDE 16

Usage

  • For Pharma-company: Outsource a protein

folding problem to cloud without sharing amino acid sequence private data

  • Can be use with Mapreduce or Spark

W-Sunday, October 28, 12

slide-17
SLIDE 17

Comparison

  • Encryption
  • [25]Gentry’s result
  • Secure multi party comm for large data.

W-Sunday, October 28, 12

slide-18
SLIDE 18

Questions ?

18

W-Sunday, October 28, 12