DocumentCode :
1678064
Title :
Mathematical Model and Quantitative Research Method on the Variability of Task Execution-time
Author :
Ming, Liu ; Chao, Yan
Author_Institution :
Inst. of Acoust., Beijing, China
fYear :
2012
Firstpage :
397
Lastpage :
402
Abstract :
The task execution-time is unpredictability on the modern complex digital signal processing platform. This paper proposed the mathematical model and the quantitative research method on the variability of task execution-time. On the modern complex digital signal processing platform, the task execution-time probability density can be approximated as normal distribution. This paper proposed the reliability of execution-time as an indicator to measure the probability that the variability of task execution-time will not induce to the error execution of task. The reliability of task execution-time will be increased by increasing the block-processing granularity. This paper proposed the method to design GPP-based digital signal processing platform whose task execution-time is determinism, and analyzed the variability of task execution-time on the Godson 2E processor-based PGC digital demodulation system. Experiments show that the quantitative research method is an effective method to research on the variability of task execution-time.
Keywords :
digital signal processing chips; normal distribution; probability; GPP-based digital signal processing platform design; Godson 2E processor-based PGC digital demodulation system; block-processing granularity; execution-time reliability; mathematical model; normal distribution; quantitative research method; task error execution; task execution-time is determinism; task execution-time probability density; task execution-time variability; Computer architecture; Demodulation; Digital signal processing; Equations; Pipelines; Program processors; Reliability; Godson; digital signal processing; normal distribution; task execution-time; variability;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Distributed Control and Intelligent Environmental Monitoring (CDCIEM), 2012 International Conference on
Conference_Location :
Hunan
Print_ISBN :
978-1-4673-0458-0
Type :
conf
DOI :
10.1109/CDCIEM.2012.101
Filename :
6178499
Link To Document :
بازگشت