The Transactions of the Korea Information Processing Society (한국정보처리학회논문지)

Korea Information Processing Society (한국정보처리학회)
권호 표지

The Optimal Distributed Database System Design Using the Genetic Algorithm

유전자 알고리즘을 이용한 최적의 분산 데이터베이스 시스템 설계

  • Ko, Suc-Bum (Dept.of Compute Engineering, Graduate School of Pukyong National University) ;
  • Youn, Sung-Dae (Dept.of Computer Science, Pukyong National University)
  • 고석범 (부경대학교 대학원 전자계산학과) ;
  • 윤성대 (부경대학교 전자계산학과)
  • Published : 2000.09.01
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, DDSs(Distributed Database Systems) have been implemented on V AN(V alue Added Network) as we know the amazing expansion of information network. DDS can yield significant cost and response time advantages over centralized systems for geographically distributed organizations. However, inappropriate design can result in high cost and poor response time to maintain the database at each site. In a DDS design, the main problem is how to select proper computer and how to allocate data fragment into a proper site. In this paper, we address DDS design problem of selecting the proper class of computers and the allocating data files on VAN. Also, the formulated model includes two objectives, the waited response time and the investment cost to include their relationship. Specially, the formulation of waited response time is based on M/M/1 queueing system to evaluate more precisely. GA(Genetic Algorithm), a kind of heuristic search method, is developed to search an optimal solution in the proposed design model and we show the simulation result to examine the algorithm performance.

최근에 정보네트워크 사용자의 급증에 따라 DDS(Distributed Database System)는 VAN(Value Added Network)상에서 구현되었다. DDS는 지역적으로 분산된 작업환경에서 중앙집중식 데이터베이스 구축보다 여러 측면에서 장점이 있으나 불합리한 설계는 컴퓨터 및 네트워크 자원의 비효율적 사용에 의한 비용의 증가와 데이터 유지를 위한 복잡도의 증가를 야기한다. DDS 설계시 각 사이트에서 적절한 컴퓨터를 선택하는 문제와 단편화된 데이터를 적절한 사이트에 할당하는 문제가 중요하다. VAN 상에서 컴퓨터 선택과 데이터 파일의 할당은 응답대기시간(waited response time)과 투자비용(investment cost)의 상관관계를 반드시 고려하여 결정되어야 하므로, 본 논문에서는 각 컴퓨터와 파일의 할당의 영향에 따라 두 목적함수의 상관관계를 고려한다. 특히, 응답대기 시간에 대한 보다 실제적인 평가를 위해 M/M/1 큐잉 시스템을 기초로 하여 설계한다. 제안된 설계모델은 경험적 탐색법 중의 하나인 유전자 알고리즘(Genetic Algorithm)의 적용을 통해 효율적인 해의 탐색을 시도하고 제안된 수학적 모델과 알고리즘의 성능 검토를 위해 모의실험 및 결과분석을 한다.

Keywords

References

  1. S. T. March and S.K. Rho, 'Allocating Data and Operations to Nodes in Distributed Database Design,' IEEE Trans. on Knowledge and Data Engg., Vol.7, No.2, pp.305-317, Apr. 1995 https://doi.org/10.1109/69.382299
  2. M. Ozsu and P. Valduriez, 'Principles of Distributed Database Systems', Prentice-Hall Inc., Englewood Cliffs, N.J. 1991
  3. S. Ram and R. E. Marsten, 'A Model for Database Allocation Incorporating a Concurrency Control Mechanism,' IEEE Trans. on Knowledge Data Eng., Vol.3, pp.389-395, Sep. 1991 https://doi.org/10.1109/69.91051
  4. H. K. Jain, 'A Comprehensive Model for the Design of Distributed Computer Systems,' IEEE Trans. on Software Eng., VoI.SE-13, pp.1092-1104, Oct. 1987
  5. M. Gen and R. W. Cheng, 'Genetic Algorithms and Engineering Design,' John Wiley and Sons, New York, 1997
  6. Z. Michalewicz, 'Genetic Algorithms+Data Structures = Evolution Programs,' second edition, Springer-Verlag, New York, 1994
  7. R. J. A. Buhr and C. M. 'Woodwide, Microscopic Economic Planning Models for Distributed Infor-mation Systems,' INFOR, Vol.15, No.2, 1977
  8. I. Mitrani and Sevcik, 'Evaluating the Trade-off between Centralized and Distributed Computing,' Proc. 1st Int. Conf. Distributed Computing Systems, pp.520-528, Oct. 1979
  9. W. W. Chu, 'Optimal File Allocation in Multiple Computer Systems,' IEEE Trans. Comput., Vol.C-18, pp.885-889, 1969 https://doi.org/10.1109/T-C.1969.222542
  10. S. Mahmoud and J. S. Riordan, 'Optimal Allocation of Resources in Distributed Information Networks,' ACM Trans. on Database System, Vol.1, No.1, pp.66-78, 1976 https://doi.org/10.1145/320434.320449
  11. J. P. Ignizio, D. F. Palmer, and C. Murphy, 'A Multicriteria Approach to Super System Architecture Definition,' IEEE Trans. on Comput., Vol.C-31, pp.410-418, May 1982
  12. A. Dutta and H. Jain, 'A DSS for Distributed Computer System Design in the Presence of Multip-le Conflicting Objectives,' Decision Support System, Vol.1, No.3, pp.233-246, Sept. 1985 https://doi.org/10.1016/0167-9236(85)90242-8
  13. C. Hwang and K. Yoon, 'Multiple Attribute Decision Making Methods and Applications,' Springer-Verlag, Berlin, 1981
  14. M. Gen, Y. Tsujimura and S.B. Ko, 'Allocation Strategy for Distributed Database System with Fuzzy Data Using Genetic Algorithm,' 5th European Congress on Intelligent Techniques and Soft Com-puting (EUFIT'97), Vol.1, pp.737-742, Sep. 1997
  15. S. Ko, Y. Tujumura and M. Gen, 'Data Distribution Considered Communication Flow in Network Using Genetic Algorithm,' 2nd Inter. Conf. on Knowledge-Based Intelligent Electronic Systems, Vol.2, pp, 264-271, Apr. 1998 https://doi.org/10.1109/KES.1998.725921
  16. J. Jo, S. Ko, S. Yoon, Y. Tsujimura, and M. Gen, 'Processor Selectio and Data Allocation Problem in Distributed Database System Using GA,' Inter. Conf. on APIEMS, pp.357-360, Oct. 1999
  17. D. C. Little, 'A Proof of the Queueing Formula L=${\lambda}$W,' Operation Res.,' Vol.9, pp.383-387, 1961