Structuring MPC

from The MPC Lounge

Sep 6, 2014, 12:13:21 PM

It is the second day of SCN, and it is time for the second session on MPC. Jason Perry is in front of the projector to talk about “Systematizing Secure Computation for Research and Decision Support”, a joint work with Debayan Gupta, Joan Feigenbaum and Rebecca Wright. More specifically the talk is a description of a new software project to categorize MPC protocols. The software includes a database compiled by the authors of different MPC protocols previously published, along with a GUI tool to make the database accessible for users. More specifically, all protocols are projected onto around 22 discrete axis, describing the different features of MPC protocols, going form weak to strong. For example, amount of corruptions; the GUI has a bar where you can adjust if you just need a protocol secure against corruption of one third of the parties, less than half, or everyone except one. The same goes for maliciousness; you can adjust whether you need a protocol that is just secure against semi-honest corruptions, covert corruptions or full-blown malicious corruptions. Unfortunately, not all aspects of protocols are easily projected onto an axis, for example security assumptions. This is solved in different ways, for the security assumptions these are handled by two axis; one binary choosing wether you want general assumptions, or if it ok to assume specific assumptions. The other is a “level indicator”, going from “none” to LWE and company, through one-way functions and trapdoor permutations. Needless to say, the categorization of MPC protocols if far from easy, and perhaps it is even harder to decipher all papers to uncovers the gritty details in order to project the protocols onto the axis of the database. This challenge is one of the reasons the authors decided to undertake the work: It is a very complex and time-consuming task to find the MPC protocols that actually fit a given setting. Their hope is that their work will make this easier and in particular increase the adoption of MPC in the “real world”. Furthermore, their project also supports impossibility results, so should you desire a protocol with features that are theoretically impossible to achieve, the software will tell you so, and thus save you from roaming the Internet looking for something that can never be found.
Currently Jason and his co-authors have 190 papers in their database and their GUI is called SysSC-UI, and is open source. As future work Jason hopes that the project can move towards a community driven model where authors themselves might submit their protocols. Furthermore, finding other and easier ways to visualize protocols would also be desirable as future work.