Open AccessFlexible Mixed-Criticality Scheduling with Dynamic Slack Management
Author(s) -
Xinyang Dong,
Gang Chen,
Mingsong Lv,
Weiguang Pang,
Wang Yi
Publication year - 2021
Publication title -
journal of circuits, systems, and computers
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.253
H-Index - 31
eISSN - 1793-6454
pISSN - 0218-1266
DOI - 10.1142/s0218126621503060
Subject(s) - computer science , mixed criticality , criticality , scheduling (production processes) , distributed computing , dynamic priority scheduling , quality of service , task (project management) , fixed priority pre emptive scheduling , execution time , worst case execution time , real time computing , rate monotonic scheduling , mathematical optimization , computer network , engineering , physics , systems engineering , nuclear physics , mathematics
Mixed-criticality (MC) system has attracted a lot of research attention in the past few years for its resource efficiency. Recent work attempted to provide a new MC model, the so-called Flexible Mixed-Criticality (FMC) task model, to relax the pessimistic assumptions in classic MC scheduling. However, in FMC, the behavior of MC tasks is still analyzed in offline stage. The run-time behavior such as dynamic slack has not yet been studied in FMC scheduling framework. In this paper, we present a utilization-based slack scheduling framework for FMC tasks. In particular, we monitor task execution on run time and collect dynamic slacks generated by task early completion. And these slacks can then be used either by high-criticality tasks to reduce mode-switches, or by low-criticality tasks so that less suspensions are triggered with more execution time, and thus quality of service is improved. We evaluate our approach with extensive simulations, and experiment results demonstrate the effectiveness of the proposed approaches.