Optimization of multi-digit Multiplication based on FFT in Parallel computational model

Authors

  • Андрій Миколайович Терещенко Senior software developer LLC «SimCorp – Ukraine»

DOI:

https://doi.org/10.18372/2410-7840.16.7533

Keywords:

parallel computational model, asymmetric cryptography, multidigit arithmetic, multidigit multiplication, DFT, FFT

Abstract

It is considered the operation of multi-digit multiplicationfor parallel computational model, that has biggest influenceon performance of asymmetric cryptographic computer systems. It is given modification of N-digit multiplicationalgorithm based on FFT and DTF`s coefficientspreviously computed. New algorithm operates with multidigitsof the length of N, contrary to standard algorithmthat uses multi-digits of the length of 2N. Algorithm reducesin two times the number of used parallel processorskeeping the same computational complexity of each processorin comparison with standard algorithm. Givenalgorithm is also efficient in sequential computationalmodel.

Author Biography

Андрій Миколайович Терещенко, Senior software developer LLC «SimCorp – Ukraine»

Ph.D in physics and mathematics

References

. Zadiraka V.K. (1985) «About requirements for

parallel data processing in FFT processor» Algorithm

and computer software optimization, Kyiv.: V.M.

Glushkov Institute of Cybernetics of NAS of

Ukraine, P. 3-10.

. Zadiraka V., Oleksyuk O. (2003) «Computer

multidigit arithmetic: Naukove vidanya», Kyiv, 263 p.

. Tereshchenko A.N. (2008) «Multiplication of big

numbers of the length of N with computing only

DTF of the length of N», Computer mathematics,

№ 1, P. 122-130.

. Tereshchenko A.N., Zadiraka V.K. (2012)

«Optimization of big numbers of the length of N

multiplication based on DTF of the length of N»,

Programming problems, № 4, P. 116-130.

Published

2015-02-03

Issue

Section

Articles