layout | sidebar | title | author | room | date | time | abstract | intropapers | advancedpapers | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
student |
true |
Garbled Circuits |
Per Hallgren |
EDIT 3364 |
2017/04/18 |
15:30-17:30 |
per.pdf |
Garbled Circuits is a technique used to compute a function based on inputs from multiple parties without disclosing any data that correlated to the user's input, except for the output of the function. The technique has been around and under development for about 30 years, but has only recently seen enough performance improvements to be feasible for real applications. We detail the main ideas of the technique, and highlight two recent developments that lead to performance improvements. The first such improvement, TinyGarble, shows how to better utilize existing optimization tools used in computer architecture. The second improvement, called GraphSC, lends support to secure computation of graph algorithms. |
|
|