NOZAKI Yusuke

Starting year 2023

Meijo University
Faculty of Information Engineering
Assistant Professor

Research fields

Information Security

Research Interests

Lightweight Cipher
Side-Channel Analysis
Physically Unclonable Function
AI Security
Energy Harvesting

Professional Memberships

The Institue of Electrical and Electronics Engineers(IEEE)
The Institute of Electronics, Information and Communication Engineers(EiC)
Information Processing Society of Japan(IPS)
Japan Society for Fuzzy Theory and Intelligent Informatics(SOFT)

Main research topics

We have been studied security in hardware. To deal with information safely and securely, it is important to ensure security in the hardware which supports the information system. In Society 5.0 which is promoted by the Japanese government, artificial intelligence (AI) and internet of things (IoT) technologies are closely related in real and virtual space. We have been studying side-channel analysis and countermeasures for cryptographic circuit, AI systems, and electronic fingerprints used in Society 5.0. Side-channel analysis is an analysis which estimates secret information using physical information such as processing time, power consumption, and electromagnetic waves during hardware operation. The study of side-channel analysis of cryptographic hardware targets lightweight ciphers. Lightweight cipher is a cryptographic technique which can be used in small area, low power, and low latency, and it is expected to be used in small devices used in IoT. We have developed power, electromagnetic, and fault analysis methods and its countermeasures for lightweight ciphers in various fields.
The study of AI security focuses on attacks that misclassify AI recognition, extract AI models, and poison AI models. We have developed an attack method to misclassify AI recognition using backdoor circuits and its countermeasures, and a model extraction attack using side-channel information and its countermeasure.
Physically Unclonable Function (PUF) is a technology to authenticate electronic devices which uses semiconductor manufacturing variations as unique ID for each device. We have developed PUF using edge AI mechanisms, PUF for IoT devices using lightweight ciphers, and PUF technology using energy harvesting, which is expected to be used in the IoT.

 

Representative papers

Y. Nozaki, S. Takemoto, Y. Ikezaki, and M. Yoshikawa, “RSM Based Tamper Resistant Implementation of Midori128 and its Evaluation,” IEICE TRANSACTIONS on Communications, vol. J106-B, no. 3, pp. 103–111, Mar. 2023.　(Japanease)

Y. Nozaki, K. Shibagaki, S. Takemoto, and M. Yoshikawa, “AI Hardware Oriented Neural Network Physical Unclonable Function and its Evaluation,” IEEJ Transactions on Electronics, Information and Systems, vol. 140, no. 7, pp. 689–696, Jul. 2020.　(Japanease)

Research URL

researchmap　https://researchmap.jp/yusuke_nozaki?lang=ja

Interview

  • No interview

News

  • No related news

Back