ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Accelerating Sensor Software Performance on Edge Devices through Just-in-Time Compilation
Cited 0 time in scopus Share share facebook twitter linkedin kakaostory
Authors
Jongwoo Choi, Wan-Ki Park, Jinsoo Han, Sangbyeong An, Sihyun Park, Chool Yu
Issue Date
2024-10
Citation
International Conference on Information and Communication Technology Convergence (ICTC) 2024, pp.2232-2233
Publisher
IEEE
Language
English
Type
Conference Paper
DOI
https://dx.doi.org/10.1109/ICTC62082.2024.10827129
Abstract
This study introduces a software-defined sensor (SDS), a software system that processes data from pre-installed hardware sensors to extract new, valuable information. An SDS typically involves either a decentralized edge computing device or a centralized computing server. The use of an edge computing device can often be problematic due to its low-performance microcontroller, which may prolong the SDS's response time. This study presents a method to enhance the response time of an SDS by applying just-in-time (JIT) compilation to its sensor software. The sensor software, developed in Python, is compiled using a JIT compiler instead of being interpreted. Experimental results demonstrate that JIT compilation nearly doubles the computation speed of the sensor software.
KSP Keywords
Edge Computing, Edge devices, JIT Compilation, JIT compiler, Just-in-Time Compilation, Software performance, Software-Defined, response time, software system