지능형 인공지능 시스템을 위한 동적 부분 업데이트 기술 기반 고속 간헐적 임베디드 소프트웨어 관리

Fast and Intermittent Embedded Software Management based on Dynamic Partial Update Techniques for Intelligent AI-Driven Systems

초록

As embedded systems evolve, embedded system updates becomes more important. In particular, over-the-air(OTA) technology is widely used via wireless for updating automotive control algorithm. Applying OTA technology for updating mission-critical control software has several problems due to characteristics of wireless communication. First, the on-chip software update on driving automobiles may be paused during the update due to various intervention or disconnection of network. Second, the transmission data may be changed due to poor update channel status. Third, the update file may be eavesdropped. In this paper, we propose a fast and secure update protocol that is intermittent and can cope with channel errors. In addition, partial update and runtime update is possible by designed to be applied to systems using dynamic library. The experimental results for dynamically linked software show our approach is effective for seamless software update under intermittent environment.

키워드

over-the-air update; intermittent update; partial update; runtime update; dynamic library; over-the-air 업데이트; 간헐적 업데이트; 부분적 업데이트; 런타임 업데이트; 동적 라이브러리
제목
지능형 인공지능 시스템을 위한 동적 부분 업데이트 기술 기반 고속 간헐적 임베디드 소프트웨어 관리
제목 (타언어)
Fast and Intermittent Embedded Software Management based on Dynamic Partial Update Techniques for Intelligent AI-Driven Systems
저자
이장훈; 박대진
DOI
10.6109/jkiice.2025.29.3.420
발행일
2025-03
유형
Y
저널명
한국정보통신학회논문지
권
29
호
3
페이지
420 ~ 431