[1]姚爱红,张国印,关 琳.基于动态可重构FPGA的自演化硬件概述[J].智能系统学报,2008,3(5):436-442.
YAO Ai-hong,ZHANG Guo-yin,GUAN L in.A survey of dynam ically and partially reconf igurable FPGA-based self-evolvable hardware[J].CAAI Transactions on Intelligent Systems,2008,3(5):436-442.
点击复制
YAO Ai-hong,ZHANG Guo-yin,GUAN L in.A survey of dynam ically and partially reconf igurable FPGA-based self-evolvable hardware[J].CAAI Transactions on Intelligent Systems,2008,3(5):436-442.
基于动态可重构FPGA的自演化硬件概述
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
3
期数:
2008年第5期
页码:
436-442
栏目:
综述
出版日期:
2008-10-25
- Title:
- A survey of dynam ically and partially reconf igurable FPGA-based self-evolvable hardware
- 文章编号:
- 1673-4785 (2008) 05-0436-07
- 分类号:
- TP303
- 文献标志码:
- A
- 摘要:
- 演化硬件研究如何利用遗传算法进行硬件自动设计,或者设计随外界环境变化而自适应地改变自身结构的硬件,在电子设计自动化、自主移动机器人控制器、无线传感器网络节点等领域都有潜在的应用价值. 自演化硬件是在硬件内部完成遗传操作和适应度计算,利用支持动态部分可重构的FPGA芯片上的微处理器核实现遗传算法,模拟生物群体演化过程搜索可能的电路设计并配置片上的可重构逻辑,找到最优或较优的设计结果,从而实现自适应硬件. 当电路发生故障时,自演化硬件自动搜索新的配置,利用片上冗余资源取代故障区域,从而实现自修复硬件. 介绍了基于动态部分可重构FPGA的自演化硬件的基本思想、体系结构以及研究现状,总结并提出了亟待解决的关键技术,指出高效的电路染色体编码表示与可重构逻辑配置位串之间的映射方式是当前研究的重点之一.
- Abstract:
- The objective of evolvable hardware (EHW) research is the development of automated electronic2circuit design, or a system capable of adap tive alterations to the hardware of its structure. EHW has a great variety of ap2 p lications in fields such as design automation, controllers for autonomous mobile robots, and wireless sensor net2 work nodes. Self2evolutionary hardware comp letes genetic operations and the evaluation of fitness in an on2chip mi2 crop rocessor. The on2chip microp rocessor core of the dynamically and partially reconfigurable field2p rogrammable gate array ( FPGA) runs the genetic algorithm, simulates the evolving p rocess to search for possible circuits, and then sets the on2chip reconfigurable logic so that it runs designswhich are superior to the originalman2made ones, or are op timal designs. In thisway adap tive hardware is possible. When a circuit fails, the self2evolutionary hard2 ware can automatically search for new configurations, using the abundant resources on the chip to rep lace the failed area, realizing the goal of self2repairing hardware. In this paper, the ideas behind self2EHW research, the archi2 tecture of self2EHW and the latest p rogress in the area are studied. The p rimary p roblems to be solved are then summarized. Finally, future research directions are pointed out, such as ways to establish a highly efficient map2 p ing mode between chromosomal encoding of circuits and the configuration bit strings of reconfigurable logic.
- 参考文献/References:
-
[ 1 ]YAO X, H IGU ICH I T. Promises and challenges of evolvab2 le hardware[ J ]. IEEE Trans on SystemsMan and Cybernet2 ics, 1999, 29 (1) : 87297.
?[ 2 ]BENTLEY P, GORDON T, KIM J , et al. New trends in evolutionary computation[C ] / / Proc 2001 Congress on Evo2 lutionary Computation. Seoul, South Korea, 2001: 1622 169.
[ 3 ]LOHN J , HORNBY G. Evolvable hardware: using evolu2 tionary computation to design and op timize hardware systems [ J ]. IEEE Computational Intelligence Magazine, 2006, 1 (1) : 19227.
[ 4 ] THOMPSON A, LAYZELL P, ZEBULUM R S. Exp lora2 tions in design space: unconventional electronics design through artificial evolution[ J ]. IEEE Trans on Evolutionary Computation, 1999, 3 (3) : 1672196.
[ 5 ] SUBB IAH P, RAMAMURTHY B. The study of fault toler2 ant system design using comp lete evolution hardware [ C ] / / Proc 2005 IEEE International Conference on Granular Com2 puting. Beijing, China, 2005: 6422645.
[ 6 ] STEFATOS E, ARSLAN T. An efficient fault2tolerantVLSI architecture using parallel evolvable hardware technology [ C ] / /Proc 2004 NASA /DoD Conference on Evolvable Hardware. Seattle, USA, 2004: 972103.
?[ 7 ]UPEGU IA, SANCHEZ E. Evolving hardware by dynami2 cally reconfiguring Xilinx FPGA [C ] / /Proc the International Conference on Evolvable Systems ( ICES ). Barcelona, Spain, 2005: 56265.
[ 8 ]RESANO J , MOZOS D, VERKEST D, et al. A reconfigu2 ration manager for dynamically reconfigurable hardware[ J ]. IEEE Trans on Design & Test of Computers, 2005, 22 (5) : 4522460.
[ 9 ] SAPARGAL IYEV Y, KALGANOVA T. Absolutely free ex2 trinsic evolution of passive low2pass filter [ C ] / /Canadian Conference on Electrical and Computer Engineering. Otta2 wa, Canada, 2006: 121021213.
?[ 10 ]ZHANG Y, SM ITH S, TYRRELL A. Digital circuit design using intrinsic evolvable hardware[C ] / /Proc 2004 NASA / DoD Conference on Evolvable Hardware. Krakow, Port2 land, 2004: 55262.
[ 11 ] TUFTE G, HADDOW P. Biologically2insp ired: a rule2 based self2reconfiguration of a virtex chip [ C ] / /Proc the Computational Science ( ICCS). Krakow, Poland, 2004: 124921256.
?[ 12 ]MORENO J , SANCHEZ E, CABESTANY J. An in2system routing strategy for evolvable hardware p rogrammable p lat2 forms[C ] / /Proc the 3 rd NASA /DoD. Long Beach: IEEE Computer Society Press, 2001: 1572166.
[ 13 ]EVANS J , ARSLAN T. The imp lementation of an evolvab2 le hardware system for real time image registration on a system2on2chip p latform[C ] / /Proc 2002 NASA /DoD Con2 ference on Evolvable Hardware. Alexandria, Virginia, 2002: 1422146.
[14 ]赵曙光, 杨万海. 一种用于演化硬件的染色体编码新方法[ J ]. 西安电子科技大学学报, 2000, 27 (6) : 7782 780.
ZHAO Shuguang, YANGWanhai. A new method for the chromosome rep resentation in the evolvable hardware [ J ]. Journal of Xidian University, 2000, 27 (6) : 7782780.
?[ 15 ]Xilinx Inc. Virtex2II p latform FPGA user guide[ EB /OL ]. [ 2005203224 ]. http: / /china. xilinx. com / support/docu2 mentation /user_guides/ug002. pdf.
[ 16 ]康立山, 何 魏, 陈毓屏. 用函数型可编程器件实现演化硬件[ J ]. 计算机学报, 1999, 22 (7) : 7812784.
KANG L ishan, HEWei, CHEN Yup ing. Evolvable hard2 ware realized with function type p rogrammable device [ J ]. Chinese Journal of Computers, 1999, 22 (7) : 7812784.
[ 17 ] L IU H, M ILLER J , TYRRELL A. Intrinsic evolvable hardware imp lementation of a robust biological develop2 ment model for digital systems [ C ] / /Proc 2005 NASA / DoD Conference on Evolvable Hardware. Washington D C, 2005: 87292.
?[ 18 ]LEV I D, GUCCIONE S. Evolvable genetic FPGA: evol2 ving stable circuits on mainstream FPGA devices Hardware [ C ] / /Proc the 1 st NASA /DoD. Pasadena, USA, 1999: 12217.
?[ 19 ]Xilinx Inc. Two flows for partial reconfiguration: Module based or difference based [ EB /OL ]. [ 2004209211 ]. ht2 tp: / /www. xilinx. com / support/documentation / app lication _notes/xapp290. pdf.
[ 20 ] SANTIN I C, AMARAL J , PACHECOM, et al. Evolution2 ary analog circuit design on a p rogrammable analog multi2 p lexer array [ C ] / /Proc IEEE International Conference on Field2Programmable Technology ( FPT2002 ). HongKong, China, 2002: 189-196.
[ 21 ] SIPPER M, SANCHEZ E, MANGE D, et al. The POE Model of bio2insp ired hardware systems: a short introduc2 tion[C ] / /Proc the Second Annual Conference of Genetic Programming. Stanford, USA, 1997: 5102511.
[ 22 ]OREIFEJ R, AlHADDAD R, HENG T, et al. Layered app roach to intrinsic evolvable hardware using direct bits2 tream manipulation of virtex II p ro devices [ C ] / / Interna2 tional Conference on Field Programmable Logic and App li2 cations( FPL 2007). Amsterdam, 2007: 2992304.
[ 23 ]NASA AMES Research Center. AMES technology capabili2 ties and facilities [ EB /OL ]. [ 2007205229 ]. http: / /www. nasa. gov/ centers/ ames/ research / technology2onepagers/ evolvable_systems. html.
[ 24 ] KAJ ITAN I I, HOSH INO T, IWATA M, et al. Variable length chromosome GA for evolvable hardware [ C ] / /Proc IEEE International Conference on Evolutionary Computa2 tion. Nagoya, Japan, 1996: 4432447.
?[ 25 ] SAKANASH I H, IWATA M, H IGUCH I T. Evolvable hardware for lossless comp ression of very high resolution bi2level images [ C ] / /Proc IEE Computers and Digital Techniques. London, UK, 2004: 2772286.
[ 26 ] LEHRE P, HADDOW P. Developmental mapp ings and phenotyp ic comp lexity [ C ] / /Proc the IEEE Congress on Evolutionary Computation (CEC). Portland, USA, 2003: 62268.
[ 27 ] GARV IE M, THOMPSON A. Evolution of combinatorial and sequential on2line self2diagnosing hardware[C ] / /Proc the 2003 NASA /DoD Conference on Evolvable Hardware. Chicago: IEEE Computer Society Press, 2003: 1672173.
[ 28 ]GAR IS H, KORKIN M, GERS F, et al. ATR’s artificial brain (CAM2brain) p roject: a samp le of what individual CoDi21Bitmodel evolved neural netmodules can do[C ] / / Proc the IEEE Congress on Evolutionary Computation (CEC). Washington,DC,USA, 1999: 197921987.
[ 29 ]AROSTEGU IM, SANCHEZ E, CABESTANY J. An in2 system routing strategy for evolvable hardware p rogramma2 ble p latforms [ C ] / /Proc The Third NASA /DoD. Long Beach, 2001: 1572166.
[ 30 ]GORDON T, BENTLEY P. Development brings scalability to hardware evolution [ C ] / /Proc the 2005 NASA /DoD Conference on Evolvable hardware. Washington,DC,USA, 2005: 2722279.
[ 31 ]GALLAGHER J , V IGRAHAM S, KRAMER G. A family of compact genetic algorithms for intrinsic evolvable hard2 ware [ J ]. IEEE Trans on Evolutionary Computation, 2004, 8 (2) : 1112126.
备注/Memo
收稿日期: 2007-06-19.
作者简介:
姚爱红,女, 1972 年生,工学博士, 副教授,中国电子学会高级会员. 主要研究方向为演化计算、嵌入式系统设计、多媒体传感器网络. 发表论文11 篇,其中被EI、ISTP检索收录6篇,出版著作和译著3部.
张国印,男, 1962年生,工学博士, 教授,博士生导师. 中国计算机学会计算机教育专委会委员、中国电子学会嵌入式系统专家委员会委员、黑龙江省计算机学会理事. 主要研究方向为嵌入式系统、智能控制与智能机器人、网络与信息安全. 获部级科技进步一等奖1项、二等奖1项,获第四届黑龙江省青年科技奖,获省高校教学成果一等奖和二等奖各1项. 发表论文70多篇,多篇被SCI、EI、ISTP检索,出版著作和译著4部.
关 琳,女, 1983 年生,硕士研究生,主要研究方向为演化计算、可重构计算系统.
通信作者:姚爱红. E-mail: yaoaihong@hrbeu. edu. cn
更新日期/Last Update:
2009-05-18