MPC with Low Bottleneck-Complexity: Information-Theoretic Security and More
| Autor: | Keller, Hannah, Orlandi, Claudio, Paskin-Cherniavsky, Anat, Ravi, Divya |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2023 |
| Předmět: | |
| DOI: | 10.4230/lipics.itc.2023.11 |
| Popis: | The bottleneck-complexity (BC) of secure multiparty computation (MPC) protocols is a measure of the maximum number of bits which are sent and received by any party in protocol. As the name suggests, the goal of studying BC-efficient protocols is to increase overall efficiency by making sure that the workload in the protocol is somehow "amortized" by the protocol participants. Orlandi et al. [Orlandi et al., 2022] initiated the study of BC-efficient protocols from simple assumptions in the correlated randomness model and for semi-honest adversaries. In this work, we extend the study of [Orlandi et al., 2022] in two primary directions: (a) to a larger and more general class of functions and (b) to the information-theoretic setting. In particular, we offer semi-honest secure protocols for the useful function classes of abelian programs, "read-k" non-abelian programs, and "read-k" generalized formulas. Our constructions use a novel abstraction, called incremental function secret-sharing (IFSS), that can be instantiated with unconditional security or from one-way functions (with different efficiency trade-offs). LIPIcs, Vol. 267, 4th Conference on Information-Theoretic Cryptography (ITC 2023), pages 11:1-11:22 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|