2016 年 11 月 Journal on Communications November 2016
2016215-1
第 37 卷第 11 期 通 信 学 报 Vol.37
No.11
基于动态自适应离散粒子群算法的 3D NoC 低功耗映射方法
刘勤让,戴启华,沈剑良,赵博
(国家数字交换系统工程技术研究中心,河南 郑州 450000)
摘 要:相对于 2D NoC,3D NoC 具有更好的集成度和系统性能,是解决低功耗映射的一个可靠途径。在传统粒
子群算法(PSOA, particle swarm optimization algorithm)的基础上,提出了一种动态自适应离散粒子群算法
(DADPSOA, dynamic adaptive discrete particle swarm optimization algorithm)。该算法基于早熟收敛程度和个体适应
度值变化动态调整参数
ω
,不断靠近最优解;同时对粒子进行合理的解构造,减小了算法时间复杂度。仿真结果
表明,与随机映射、遗传算法(GA, genetic algorithm)、PSOA 和动态蚁群算法(DACA, dynamic ant colony algorithm)
相比,DADPSOA 可以缩短执行时间,减小映射结果通信功耗;在面向任务图映射的时候,其通信功耗下降。
关键词:3D NoC;低功耗映射;解构造;自适应离散粒子群算法
中图分类号:TP393.03 文献标识码:A
Dynamic adaptive discrete particle swarm optimization algorithm
based method on low-power mapping in network-on-chip
LIU Qin-rang, DAI Qi-hua, SHEN Jian-liang, ZHAO Bo
(National Digital Switching System Engineering & Research Center, Zhengzhou 450000, China)
Abstract: Compared to 2D NoC, 3D NoC has better integrated density and system performance, which was a reliable
method to solve the problem about low-power mapping. On the basis of the traditional particle swarm optimization algo-
rithm (PSOA), a dynamic adaptive discrete particle swarm optimization algorithm (DADPSOA) was proposed . Parame-
ter in this algorithm was adjusted dynamically based on the degree of early convergence and the charge of individual adap-
tive value to approach the optimal solution. At the same time, the reasonable structure of the particles was made aiming at
reducing the time complexity of this algorithm. Experimental results show that comparing with the random mapping, genetic
algorithm (GA), PSOA and dynamic ant colony algorithm (DACA), DADPSOA can save the execution time, reduce the
communication power consumption of mapping results. The power consumption of the task graph is reduced.
Key words: 3D NoC, low-power mapping, deconstruction, adaptive discrete particle swarm optimization algorithm
1 引言
随着 CMOS(complementary metal oxide semi-
conductor)技术的发展,单个芯片上集成的晶体管数
量越来越多。传统总线架构已经无法满足日益增长
的数据需求
[1,2]
。NoC (network-on-chip)凭借其可扩
展架构和并行通信的特点在一段时间内缓解数据
传输物理限制
[3,4]
。但随着芯片集成度的持续提高,
2D NoC Mesh 物理链路拥塞、传输延迟等问题越发
突出。而 3D IC(integrated circuit) 技术的逐渐成熟,
使复杂度高、通信量大的任务图能够从 2D Mesh 映
射到 3D Mesh 上,促成了 3D NoC 的产生,增强了
映射平台处理多样化任务图的能力
[5,6]
。相对于 2D
Mesh,3D NoC Mesh 具有更好的集成度和性能,是
收稿日期:2015-11-02;修回日期:2016-11-08
基金项目:国家高技术研究发展计划(“863”计划)基金资助项目(No. 2014AA01A704);国家自然科学基金创新群体基金
资助项目(No. 61521003);国家自然科学基金面上基金资助项目(No. 61572520)
Foundation Items: The National High Technology Research and Development Program of China (863 program) (No. 2014AA01A704),
The Innovation Group Program Project of National Natural Science Foundation of China (No. 61521003), The General Program
of National Natural Science Foundation of China (No. 61572520)
doi:10.11959/j.issn.1000-436x.2016215