Abstract
We describe, and implement, a maliciously secure protocol for two-party computation in a parallel computational model. Our protocol is based on Yao's garbled circuit and an efficient OT extension. The implementation is done using CUDA and yields fast results for maliciously secure two-party computation in a financially feasible and practical setting by using a consumer grade CPU and GPU. Our protocol further uses some novel constructions in order to combine garbled circuits and an OT extension in a parallel and maliciously secure setting. © 2013 Springer-Verlag.
Cite
CITATION STYLE
Frederiksen, T. K., & Nielsen, J. B. (2013). Fast and maliciously secure two-party computation using the GPU. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7954 LNCS, pp. 339–356). https://doi.org/10.1007/978-3-642-38980-1_21
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
