Secure Multiparty Computation - a technology with promises for privacy
2 likes649 views
The document discusses secure multiparty computation (SMC), emphasizing its potential for privacy and secure data handling in various applications, including agricultural trade and healthcare. It presents the benefits and drawbacks of SMC, stressing the importance of trust models in realizing its business potential. Additionally, it offers examples of SMC in practice, such as private database joins and encrypted health data retrieval.
![Using the right trust modelThe SIMAP Case: sugar beets[Bogetoft et al., Financial Crypto ‘09]Farmingcontrolled by EU-quotas (likeotheragriculturalproducts,fish, milk etc.)](https://image.slidesharecdn.com/20090924thenetwillnotforget-090920151039-phpapp01/85/Secure-Multiparty-Computation-a-technology-with-promises-for-privacy-9-320.jpg)
Secure Multiparty Computation - a technology with promises for privacy
- 2. Secure Multiparty Computation – a technology with promises for privacyJakob I. PagterCentre for IT SecurityThe Alexandra Institute Ltd.
- 4. The planWhat is SMC
- 5. Sugar Beets
- 6. Examples of SMC as a PETWhat is SMCThe ability to work on encrypted dataBenefits:Strong security
- 8. Audits
- 9. Using the right trust modelThe SIMAP Case: sugar beets[Bogetoft et al., Financial Crypto ‘09]Farmingcontrolled by EU-quotas (likeotheragriculturalproducts,fish, milk etc.)
- 10. Before 2007/2008 only bilateral tradetroughpaperads etc.
- 11. One of threefactoriesclosed and guaranteedpriceslowered => many farmerswanted to selltheirquota
- 13. Increasetrade up to 400%
- 14. Ensurealmostfullusage of national quota (lessthan 50% withoutreallocation)Trust problemHowcan the business potential berealised?!
- 15. Public-key cryptoSender and receiver each has a key: one for locking and one for unlocking
- 16. Ex:key-padeverybodycanlock it (encrypt) but only the holder of the matchingkey kan open the keypad (decrypt)Shallow Confidentialitypkbid1Epk(bid1)DecryptComputebid2Epk(bid2)Epk(bid3)bid3AuctioneerEpk(bid4)bid4Epk(bid5)bid5price
- 18. A privacy perspectiveHowcaninterestingstuffbe done withoutrevealing sensitive info
- 19. PET example 1: private DB “joins”Insurance companyNational health registerDesirable outcome
- 20. PET example 2: private information retrieval Blood sampleResult encrypted and sharedAnalysisSMCQueryresultQUERYSMCOnly the patient knows query and resultRESULTSMCSMCAnonymous computation
- 21. Further PET examples and ideasPrivate search
- 22. Storage
- 24. Personal data in the cloud
- 25. …
Editor's Notes
- #6: More on trust models later