본문 바로가기
  • Home

Task Scheduling Scheme for Maintaining Freshness in Digital Twin Systems

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2026, 31(2), pp.183~189
  • DOI : 10.9708/jksci.2026.31.02.183
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : January 12, 2026
  • Accepted : February 5, 2026
  • Published : February 27, 2026

Taehun Yang 1 Cheonyong Kim 2

1국립경국대학교
2국립한국교통대학교

Accredited

ABSTRACT

In digital twin systems, maintaining a high level of freshness for cyber twins is essential for accurate inference and real‑time decision making. However, update tasks and inference tasks are processed under limited computing resources, causing a trade-off between task throughput and digital twin freshness. Existing scheduling methods prioritize task throughput, resulting in degraded digital twin freshness under a high traffic arrival rate. To address this issue, we propose an expected-freshness–based scheduling scheme that dynamically determines the execution timing of update tasks by predicting the expected number of inference tasks within each update interval. By estimating the maximum achievable freshness and scheduling update tasks when degradation is expected, the proposed scheme ensures the required digital twin freshness level. Simulation results show that our scheme maintains digital twin freshness even under increasing task arrival rates, while incurring only moderate task rejection. The proposed scheduling scheme is suitable for applications where freshness is more critical than task throughput.

Citation status

* References for papers published after 2024 are currently being built.