diFlow

Sustainable Decentralized Secure Computation

Towards Building a Secure Multi-party Computation Infrastructure for Everyone

Our Mission

Secure multi-party computation (MPC) is a cryptographic technique that allows mutually distrusting parties to jointly evaluate a function on their private inputs without revealing anything beyond the output. MPC facilitates collaborative utilizations of confidential information held by different individuals or companies. This project aims to solve the issues related to the need for trust in centralized MPC providers and operational difficulties in the absence of such providers. To achieve these aims, we will establish a theoretical framework for efficient MPC in a fully decentralized setting while also developing an infrastructure that can be operated in a sustainable manner by incorporating users’ incentives. Our goal is to thereby provide a sustainable decentralized secure computation infrastructure.

Efficient MPC Protocols

There are many efficient MPC protocols for a small number of parties, e.g., three-party protocols. This is suitable for MPC system based on centralized providers, who perform secure computation based on e.g. three servers. We aim to construct MPC protocols that can be run efficiently for any number of parties and without central servers. We refer to this part of the project as dFlow for "Data Flow".

Incentive Infrastructures

MPC systems with centralized providers crucially rely on the providers for their operation. We aim to provide an MPC infrastructure which can be run in a decentralized manner, i.e., without any centralized providers. Towards this, we will construct a payment infrastructure that will incorporate users' (financial) incentives. We refer to this part of the project as iFlow for "Incentive Flow".

MPC + Incentives

We will finally combine MPC with incentive payments to obtain a sustainable decentralized secure computation infrastructure. A challenge in combining both is to establish fairness between dFlow and iFlow. We call this overall system as diFlow.

Crypto + Formal Methods

Our team comprises researchers in various areas: theoretical cryptography, formal verification, hardware security, and more. We aim to establish systems/protocols that are provably secure cryptographically while their functional correctness is also verifiable using formal methods. We will also seek alternative solutions based on hardware assistance, if applicable.

News / Logs

2023-11-30. Add the publication list page. Update the project member list.
2023-02-16. This web page is launched.
2023-01-16. The third full project meeting.
2022-11-29. The second full project meeting.
2022-11-10. Our research paper titled "Secure Parallel Computation on Privately Partitioned Data and Applications" was presented at the 29th ACM Conference on Computer and Communications Security (ACM CCS 2022).
2022-10-22. Full kick-off meeting for the project among (almost) all the members.
2022-10-01. The project started.

About This Project

This project is a JST CREST project within the "Society 5.0 System Software" research area (the official full name of the research area is "Creation of System Software for Society 5.0 by Integrating Fundamental Theories and System Platform Technologies"). Out project started from October 2022 and is planned to run for five and a half years.

Contact

For any queries, please contact the following e-mail address.

[mail: n.attrapadung (at-mark) aist.go.jp]

Page updated

Google Sites

Report abuse