多核阵列的任务调度技术研究-计算机科学.PDFVIP

Vo1. 44 No.8 第 44 卷第8 期 计算机科学 2017 年 8 月 COMPUTER SCIENCE Aug.2017 多核阵列的任务调度技术研究 陈亦欧吕信科凌翔 (电子科技大学通信抗干扰技术国家级重点实验室 成都 611731) 摘 要 随着信号处理的复杂度的增加，多核并行架构成为数字信号系统的有效解决方案。主要研究了面向数字信 号处理系统的元线多核阵列的任务调度问题。从数字信号处理系统与元线多核阵列的性能和开销要求出发，以功耗、 热分布以及延时为优化目标，设计出相应的功耗、热均衡评估与延时模型，作为多目标优化算法的目标函数。同时，在 NSGA-II 算法的基础上改进拥挤策略与初始种群，并设计新的适应度函数，兼顾 3 个优化目标的性能，增加探索到更 优解的可能性。最后，在无线多核阵列平台上采用多种任务图进行仿真，验证了所提算法的有效性与优越性。 关键词 多核阵列，热均衡，任务调度，多目标优化 中图法分类号 TN4 文献标识码 A 001 10. 11896/j. issn. 1002-137X. 2017. 08. 008 New Task Scheduling Technique for Multicore Arrays CHEN Yi-ou L V Xin-ke LING Xiang (National Key Lab of Science and Technology on Comrnunications ,University of Electronic Science and Technology of China ,Chengdu 611731 ,China) Abstract As the complexity of the signal processing increases ,the multicore parallel architecture has become an effec tive solution for digital signal processing (DSP) systems. This paper proposeed a new task scheduling technique in wire less multicore systems for DSP systems. Three optimization objectives ,such as power consumption ,therm且1 distribution and latency ,were chosen to achieve the performance and cost requirements of DSP systems and wireless multicore ar rays. The respective models were designed as the objective functions of the multi-objective optimization algorithrn. Also , an improved crowded strategy and the initial population selection based on NSGA-II algorithm with new fitness func- tions were proposed to balance the performance of the three optimization objectives and increase the possibility of explo ring better solutions. Finally ,