A Low Power ECC H-matrix Optimization Method using an Ant Colony Optimization

ACO를 이용한 저전력 ECC H-매트릭스 최적화 방안

  • Lee, Dae-Yeal (Department of Electrical Electronic Engineering, Yonsei University) ;
  • Yang, Myung-Hoon (Department of Electrical Electronic Engineering, Yonsei University) ;
  • Kim, Yong-Joon (Department of Electrical Electronic Engineering, Yonsei University) ;
  • Park, Young-Kyu (Department of Electrical Electronic Engineering, Yonsei University) ;
  • Yoon, Hyun-Jun (Department of Electrical Electronic Engineering, Yonsei University) ;
  • Kang, Sung-Ho (Department of Electrical Electronic Engineering, Yonsei University)
  • 이대열 (연세대학교 전기전자공학과) ;
  • 양명훈 (연세대학교 전기전자공학과) ;
  • 김용준 (연세대학교 전기전자공학과) ;
  • 박영규 (연세대학교 전기전자공학과) ;
  • 윤현준 (연세대학교 전기전자공학과) ;
  • 강성호 (연세대학교 전기전자공학과)
  • Published : 2008.01.25

Abstract

In this paper, a method using the Ant Colony Optimization(ACO) is proposed for reducing the power consumption of memory ECC checker circuitry which provide Single-Error Correcting and Double-Error Detecting(SEC-DED). The H-matrix which is used to generate SEC-DED codes is optimized to provide the minimum switching activity with little to no impact on area or delay using the symmetric property and degrees of freedom in constructing H-matrix of Hsiao codes. Experiments demonstrate that the proposed method can provide further reduction of power consumption compared with the previous works.

본 논문에서는 Ant Colony Optimization(ACO)을 이용하여 Single-Error Correcting & Double-Error Detecting(SEC-DED)을 제공하는 메모리 ECC 체커 회로의 소비전력을 절감하는 방안을 제시한다. H-매트릭스를 통해 구현되는 SEC-DED 코드인 Hsiao 코드의 대칭성과 H-매트릭스 구성상의 높은 자유도를 이용하여 회로의 면적, 딜레이에 영향을 주지 않고 최소의 비트 트랜지션이 일어나도록 H-매트릭스를 최적화한다. 실험을 통하여 H-매트릭스의 최적화를 위한 ACO 매핑과 파라메터의 설정을 알아보고 이의 구현 결과를 랜덤 매트릭스 구성을 통한 방식 및 기존의 GA알고리즘을 이용한 최적화 방식과 비교하여 소비 전력이 기존의 방식에 비해 절감될 수 있음을 보여준다.

Keywords

References

  1. Chen, C.L and Hsiao, M.Y.: 'Error-correcting codes for semiconductor memory applications: a state of the art review', IBM J. Res. Develop., vol. 28, pp. 124-134, July 1984 https://doi.org/10.1147/rd.282.0124
  2. Mohanram, K., and Touba, N.A.: 'Input ordering in concurrecnt checkers to reduce power consumption', Proc. of IEEE Sumposium on Defect and Fault Tolerance, pp. 87-95, 2002
  3. Rossi, D., van Dijk, V.E.S, Kleihorst, R.P., Nieuwland, A.K. and Metra, C.: 'Coding scheme for low energy consumption fault-tolerant bus', Proc. of International On-Line Testing Workship, pp. 8-12, 2002
  4. Ghosh, S., Basu, S and Touba, N.A.: 'Reducing power consumption in Memory ECC Checkers', International Test Conference, pp. 1322-1331, 2004
  5. Hsiao, M.Y.: 'A class of optimal minimum odd-weight-column SEC-DED codes', IBM J. Res. Develop., vol. 14, pp. 395-401, July 1970 https://doi.org/10.1147/rd.144.0395
  6. Heesung Lee, Joonjyung Sung and Euntai Kim: 'Reducing power in error correcting code using genetic algorithm', Proc. of Transactions on Engineering, Computing and Technology, vol. 19, pp. 179-182, January 2007
  7. Dorigo, M., Maniezzo and Colorni, A. : 'The ant system : an autocatalyic optimizing process', Technical Report 91-016 Revised, Dipartimento di Elettronica, Politecnico di Milano, Italy, 1991
  8. Dorigo, M. And Gambardella, M. : 'Ant colony system : a cooperative learning approach to the traveling salesman problem', Proc. of IEEE Transactions on Evolutionary Computation, pp. 53-66, 1997 https://doi.org/10.1109/4235.585892
  9. Stutzle, T. and Hoos, H. : 'MAX-MIN Ant System', Future Generation Computer Systems, pp. 889-914, June 2000 https://doi.org/10.1016/S0167-739X(00)00043-1
  10. Bullnheimer, B., Hartl, R.F and Strauss, C. : 'A new rank-based version of the ant system : a conputational study', Proc. of Central European Journal for Operations Research and Economics, 1997
  11. Dorigo, M. and Caro, G.D.: 'Ant colony optimization: a new meta-heuristic', Proc. of Evolutionary Computation, Vol 2, pp. 1470-1477, July 1999