DocumentCode :
2201025
Title :
Research and Implement of Dynamic Hybrid Real-time Tasks Scheduling Algorithm Based on µC/OS-II
Author :
Wang Xibo ; Tong Xin ; Du Xiaozhou
Author_Institution :
Sch. of Inf. Sci. & Eng., Shenyang Univ. of Technol., Shenyang, China
fYear :
2010
fDate :
1-3 Nov. 2010
Firstpage :
177
Lastpage :
180
Abstract :
In real-time OS, one of the most primary factor that affecting real-time performance directly is scheduling strategy. μC/OS-II, which adopting a single static priority scheduling manner, could not be well-suitable for complex hybrid tasks set scheduling. Based on EDF (Earliest Deadline First) algorithm, in combination with task importance consideration, a new dynamic scheduling algorithm EDIF (Earliest Deadline &amp, Important First) of hybrid real-time tasks is proposed. According to the algorithm idea, μC/OS-II kernel has been modified, hybrid tasks dynamic scheduling has been realized. Meanwhile, the static priority scheduling strategy of the original system being preserved, static and dynamic scheduling can be switched by modifying configuration file. Theoretical analysis and experimental results indicate that the improved dynamic scheduling algorithm can guarantee important tasks against missing their deadlines while system overload, at the same time, task completing ration of non-important tasks is also at a higher level, real-time scheduling performance of μC/OS-II is improved effectively.
Keywords :
operating system kernels; real-time systems; scheduling; μC/OS-II kernel; dynamic hybrid real-time task scheduling algorithm; earliest deadline first algorithm; real-time OS; static priority scheduling strategy; µC/OS-II; dynamic scheduling; scheduling strategy; task completing ration;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Networks and Intelligent Systems (ICINIS), 2010 3rd International Conference on
Conference_Location :
Shenyang
Print_ISBN :
978-1-4244-8548-2
Electronic_ISBN :
978-0-7695-4249-2
Type :
conf
DOI :
10.1109/ICINIS.2010.103
Filename :
5693708
Link To Document :
بازگشت