Secure Massively Parallel Computation for Dishonest Majority

Fernando, Rex; Komargodski, Ilan; Liu, Yanyi; Shi, Elaine

doi:10.1007/978-3-030-64378-2_14

Cited by 5 publications

(1 citation statement)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nonetheless, a few previous works have tackled SMPC performance issues from the network's perspective. For example, Fernando et al [13] used a point-multipoint SMPC network topology where a selected party established direct links to the other parties. This setup reduces the network complexity and traffic by having all the round data aggregated by a single party, which broadcasts the aggregated (public) data back to the parties.…”

Section: Related Workmentioning

confidence: 99%

Network-Efficient Pipelining-Based Secure Multiparty Computation for Machine Learning Applications

Bautista

Akkaya

2022

2022 IEEE 47th Conference on Local Computer Networks (LCN)

View full text Add to dashboard Cite

Secure multi-party computation (SMPC) allows mutually distrusted parties to evaluate a function jointly without revealing their private inputs. This technique helps organizations collaborate on a common goal without disclosing confidential or protected data. Despite its suitability for privacy-preserving computation, SMPC suffers from network-based performance limitations. Specifically, the SMPC parties perform the techniques in rounds, where they execute a local computation and then share their round output with the other parties. This network interchange creates a bottleneck as parties need to wait until the data propagates before resuming the execution. To reduce the SMPC execution time, we propose a pipelining-like approach for each round's computation and communication by dividing the data and readjusting the execution order. Targeting deep learning applications, we propose strategies for the case of matrix multiplication, a core component of such applications. Our results on a distributed cloud deployment show a significant reduction in the SMPC execution time.

show abstract

Section: Related Workmentioning

confidence: 99%