ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술지 Deep Neural Networks Based Key Concealment Scheme
Cited 1 time in scopus Download 122 time Share share facebook twitter linkedin kakaostory
저자
김태혁, 윤택영, 최두호
발행일
202011
출처
IEEE Access, v.8, pp.204214-204225
ISSN
2169-3536
출판사
IEEE
DOI
https://dx.doi.org/10.1109/ACCESS.2020.3036650
협약과제
20HR1900, 미래컴퓨팅 환경에 대비한 계산 복잡도 기반 암호 안전성 검증 기술개발, 최두호
초록
To keep the Internet-of-things (IoT) environment secure, employing a cryptographic function to various IoT devices has become vital. An important factor to consider is how to store a cryptographic key (or passwords) securely. A popular method is to store the key in the storage protected by some hardware-based security functions. This paper presents a novel concept to conceal cryptographic keys into deep neural networks (DNNs), named DNNs-based key concealment scheme. In this scheme, a key can be concealed into a proper deep neural network model which is trained with secret input data. We demonstrate the practical applicability of our concept by presenting an instance and a use-case scenario of the DNNs-based key concealment scheme and show its correctness. To prove its robustness, two fundamental security evaluation methods are proposed for investigating the security of the instantiation. To the best of our knowledge, this is the first attempt of its kind.
KSP 제안 키워드
Deep neural network(DNN), Evaluation method, Internet of thing(IoT), IoT Devices, Security Evaluation, Use Cases, cryptographic key, hardware-based security, input data, neural network model
본 저작물은 크리에이티브 커먼즈 저작자 표시 (CC BY) 조건에 따라 이용할 수 있습니다.
저작자 표시 (CC BY)