SystemC的扩展及其在性能分析中的应用研究  

Studying on extending SystemC and its application on performance analysis

作　　者：刘金伟[1] 黄樟钦[1] 侯义斌[1] 霍思佳[1] 王晋嘉[1] 

机构地区：[1]北京工业大学计算机学院,北京100124

出　　处：《四川大学学报（自然科学版）》2009年第6期1631-1637,共7页Journal of Sichuan University(Natural Science Edition)

基　　金：国家自然科学基金重大研究计划项目(90407017);北京市自然科学基金(KZ200510005006)

摘　　要：在嵌入式系统的系统级建模和性能分析过程中,经常需要定量地分析和评估系统的平均延时、CPU占用率等性能指标.本文基于SystemC语言,针对嵌入式系统设计了一套性能分析评估的库函数,用于统计性能参数样本、计算期望值和置信区间,同时在性能指标满足要求的精度时终止仿真.并详细阐述了性能分析库函数的设计方法.以语音识别系统为例,设计了一个虚拟仿真平台,并使用该性能分析库函数进行性能分析,仿真结果表明,设计的库函数适用于嵌入式语音识别系统的性能分析,在一定程度上对SystemC的类库进行了扩展.In the process of embedded systems system-level modeling and performance analysis,the performance indicators,such as average delay,CPU occupancy rate etc.,are often need to be analyzed and evaluated.In this paper,the library functions of performance evaluation based on SystemC language have been designed.It is designed for counting the performance parameter samples,calculating expectations and confidence interval,and terminating the simulation when the performance indicators meet the accuracy requirements.This paper also introduces the design method of the library functions of performance evaluation in detail.A virtual emulation system of speech recognition system which as an example was designed,and analyzed the performance by using the library functions.It is adequate for analyzing the performance of the speech recognition system by the results of simulation,and the library functions extended SystemC language to a certain extent.

关 键 词：SystemC语言 性能分析 应用研究 PERFORMANCE analysis speech recognition system PERFORMANCE indicators PERFORMANCE evaluation 嵌入式语音识别系统 库函数 PERFORMANCE parameter 嵌入式系统设计 confidence interval 性能指标 language average delay 虚拟仿真平台 系统级建模 paper method of analyzing 

分 类 号：N55[自然科学总论]