Parallel implementations of CHAM
- 문서 유형
- 연구출판물 · 학술대회 발표논문
- 등재
- Scopus
- 저널 순위
- Scopus Q2
- 저널
- Lecture Notes in Computer Science11402 LNCS · pp.93 · 2019
- 피인용
- Scopus 3
- FWCI (Scopus)
- 0.51
- ISSN
- 0302-9743
저자
- 서화정제1저자Hansung University, Seoul, Republic of Korea
- An, KyuhwangHansung University, Seoul, Republic of Korea
- Kwon, HyeokdongHansung University, Seoul, Republic of Korea
- 박태환Pusan National University, Busan, Republic of Korea
- Hu, ZhiCentral South University, Changsha, China
- 김호원교신저자Pusan National University, Busan, Republic of Korea
초록
© Springer Nature Switzerland AG 2019.In this paper, we presented novel parallel implementations of CHAM-64/128 block cipher on modern ARM-NEON processors. In order to accelerate the performance of the implementation of CHAM-64/128 block cipher, the full specifications of ARM-NEON processors are utilized in terms of instruction set and multiple cores. First, the SIMD feature of ARM processor is fully utilized. The modern ARM processor provides 2 × 16-bit vectorized instruction. By using the instruction sets and full register files, total 4 CHAM-64/128 encryptions are performed at once in data parallel way. Second, the dedicated SIMD instruction sets, namely NEON engine, is fully exploited. The NEON engine supports 8 × 16-bit vectorized instruction over 128-bit Q registers. The 24 CHAM-64/128 encryptions are performed at once in data parallel way. Third, both ARM and NEON instruction sets are well re-ordered in interleaved way. This mixed approach hides the pipeline stalls between each instruction set. Fourth, the multiple cores are exploited to maximize the performance in thread level. Finally, we achieved the 4.2 cycles/byte for implementation of CHAM-64/128 on ARM-NEON processors. This result is competitive to the parallel implementation of LEA-128/128 and HIGHT-64/128 on same processor.
키워드
ARM-NEON processorCHAMParallel computationSoftware implementation
식별자
- Scopus EID
- 2-s2.0-85065019692



로그인
연구물 목록으로