PARALLEL IMPLEMENTATION OF MULTI-KEY FULLY HOMOMORPHIC ENCRYPTION SCHEME CHEN, DAI, KIM, SONG (CDKS) ON GPU USING CUDA
Homomorphic encryption is an encryption scheme that allows computations on encrypted data. The Cheon-Kim-Kim-Song (CKKS) scheme is one example of current fully homomorphic encryption. CKKS scheme supports addition and multiplication operations on encrypted texts, but only for texts encrypted with...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/81834 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Homomorphic encryption is an encryption scheme that allows computations on
encrypted data. The Cheon-Kim-Kim-Song (CKKS) scheme is one example of
current fully homomorphic encryption. CKKS scheme supports addition and
multiplication operations on encrypted texts, but only for texts encrypted with the
same key. The Chen-Dai-Kim-Song (CDKS) scheme is an extended variant of the
CKKS scheme that supports operations on multiple encrypted texts with different
keys. Because CDKS can be used with many keys, it is also known as Multi-Key
FHE or multi-key fully homomorphic scheme. However, the CDKS is has not been
widely used duet o its high computational resource demands. This leads to a long
execution time when the scheme is run serially. Therefore, parallelization can be
utilized to accelerate the execution time of CDKS scheme. The parallel
implementation of the CDKS scheme uses CUDA platform on NVIDIA GPUs.
Testing was conducted by measuring the speed up between serial and parallel
CDKS scheme. The results of the implementation show that the parallel
implementation gives faster execution times than the serial CDKS scheme, with a
speed up of up to 50 times. |
|---|