Advertisement

Parallel Asynchronous Team Algorithms

  • Benjamín Barán
Part of the Applied Optimization book series (APOP, volume 67)

Abstract

Abstract Solution of today large complex problems may need the combination of several different methods, algorithms and techniques in a distributed computing system with heterogeneous processors, usually interconnected through a communication network. In this context, Team Algorithm is presented as a general technique used to combine different methods and algorithms in a distributed computing system composed of different workstations, personal computers and/or parallel computers. Moreover, experimental results have proved that in many real world problems, Team Algorithms can benefit from the use of asynchronous implementations, speeding up the whole process with an important synergy effect, in a new appealing technique known as Parallel Asynchronous Team Algorithms.

The main idea behind Parallel Asynchronous Team Algorithms is very simple: to partition a large complex problem in small sub-problems that can be solved in different processors of a distributed system with well known sequential methods, combining the partial results of each sub-problem in such a way that a good global solution is finally found. Because each processor works at its own speed and eventually, with its own algorithms, an asynchronous implementation eliminates idle synchronization times speeding up the whole process. Team Algorithms have been successfully applied in the resolution of many engineering problems in which a synergetic effect may take place between different processors.

Keywords

Distributed system asynchronous implementation parallelism network algorithm combination 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    A-Teams Project Home Page, Carnegie Mellon University. URL: http://www.cs.cmu.edu/afs/cs/project/edre-22/project/ateams/WWW/.
  2. [2]
    B. Barán, A study ofparallel asynchronous team algorithms, PhD thesis, Federal Univ. of Rio de Janeiro, COPPE, CS Dept., Rio de Janeiro, Brasil, 1993 (in Portuguese). Google Scholar
  3. [3]
    B. Barán, E. Kaszkurewicz and D.M. Falcão, Team Algorithms in Distributed Load Flow Computations, IEE Proceedings on Generation, Transmission and Distribution. Vol. 142, N° 6, November 1995, pp. 583–588.CrossRefGoogle Scholar
  4. [4]
    B. Barán, E. Kaszkurewicz and A. Bhaya, Parallel Asynchronous Team Algorithms: Convergence and Performance Analysis, IEEE Transactions on Parallel and Distributed Systems, Vol.7, Number 7, July 1996, pp. 677–688.CrossRefGoogle Scholar
  5. [5]
    B. Barán and E. Chaparro, Algoritmos Asincronos combinados en un Ambiente Heterogéneo de Red, XXIII Conferencia Latinoamericana de Informática, Valparaiso, Chile, 1997 (in Spanish). Google Scholar
  6. [6]
    B. Barán, E. Chaparro and N. Cáceres, A-Teams en la Optimización del Caudal Turbinado de una Represa Hidroeléctrica, Conferencia Iberoamericana de Inteligencia Artificial IBERAMIA-98, Lisbon, Portugal, 1998 (in Spanish). Google Scholar
  7. [7]
    B. Barán, E. Chaparro and N. Cáceres, Hydroelectric optimization using A-Teams, International Conference in Electrical Power System Operation and Management (EPSOM-98), Zurich, Switzerland, 1998.Google Scholar
  8. [8]
    B. Barán and F. Laufer, Topological Optimization of Reliable Networks using A-Teams, Symposium: Architecture, Tools and Algorithm for Networks, Parallel and Distributed Systems, Internacional Conference Systemics, Cybernetics and Informatics SCI’99, Orlando — Florida, United States of America, 1999.Google Scholar
  9. [9]
    D. P. Bertsekas and J. N. Tsitsiklis, Parallel and distributed computation — numerical methods, Prentice Hall, Englewood Cliffs, New Jersey, 1989.zbMATHGoogle Scholar
  10. [10]
    A. Bhaya, E. Kaszkurewicz, and F. C. Mota, Asynchronous block-iterative methods for almost-linear equations, Lin. Algebra Appl., 1991, pp. 487–508.Google Scholar
  11. [11]
    C. J. Coulbourn, Reliability Issues in Telecommunication Network Planning, University of Vermont. URL: http://www.emba.uvm.edu/colbourn.
  12. [12]
    C. J. Coulbourn, The Combinatorics of Network Reliability, Oxford Univ. Press, 1987.Google Scholar
  13. [13]
    J. Calvet and A. Titli, Overlapping vs. partitioning in block-iteration methods: application in large-scale system, Automatica 25,1989, pp. 137–145.MathSciNetzbMATHCrossRefGoogle Scholar
  14. [14]
    Y. P. Dusonchet, S. N. Talukdar, and H. E. Sinnot, Load flows using a combination of point Jacobi and Newton’s methods, IEEE Trans. Power Apparatus and Systems, PAS-90, 1971, pp. 941–949.Google Scholar
  15. [15]
    D.M. Falcão, A. L. B. Bomfim, C. R. R. Dornellas, and G. N. Taranto, Genetics Algorithms in Power Systems Optimization, V SEPOE, Recife — Brazil, 1996.Google Scholar
  16. [16]
    M.R. Garey and D. S. Johnson, Computers and Intractability: A Guide to the Theory of NP-Completeness, V SEPOE, Recife — Brazil, 1996.Google Scholar
  17. [17]
    D.E. Goldberg, Genetic Algorithm in Search, Optimization and Machine Learning, San Francisco, Freeman, 1979.Google Scholar
  18. [18]
    M. Ikeda and D. D. Iljak, Overlapping decomposition, expansions and contractions of dynamic systems, Large Scale Systems, Vol. 1, 1980, pp. 29–38.MathSciNetzbMATHGoogle Scholar
  19. [19]
    R. H. Jan, Design of reliable networks, Comput. Oper. Res. Vol 20, 1993, pp. 29–38.MathSciNetGoogle Scholar
  20. [20]
    E. Kaszkurewicz, A. Bhaya, and D. D. Siljak, On the convergence of parallel asynchronous block-iterative computations, Lin. Algebra Appl., 1990, pp. 139–160.Google Scholar
  21. [21]
    M. Mejia and E. Cantúi, DGENESIS: Software para la Ejecución de Algoritmos Genéticos Distribuidos, XX Conferencia Latinoamericana de Informática , Mexico City, Mexico, 1994 (in Spanish). Google Scholar
  22. [22]
    N. H. Penny, Blackboard systems: the blackboard model ofproblem solving and the evolution ofblackboard architectures, The AI Magazine, 1986.Google Scholar
  23. [23]
    W. H. Press, B. P. Flannery, S. A. Teukolsky, and W. T. Vetterling, Numerical Recipes in C: The Art of Scientific Computing, Cambridge University Press, 1988.zbMATHGoogle Scholar
  24. [24]
    P. S. Souza and S. N. Talukdar, Genetics Algorithms in Asynchronous Teams, ICGA — 91, San Diego — California, 1991, pp. 392–397.Google Scholar
  25. [25]
    M. E. Sezer and D. D. Siljak, Nested epsilon decompositions of linear systems: weakly coupled and overlapping blocks, SIAM J. Matrix Anal. Appl., 12 (1991), pp. 521–533.MathSciNetzbMATHCrossRefGoogle Scholar
  26. [26]
    P. S. Souza and S. N. Talukdar, Genetic algorithms in asynchronous teams, in Proc. Fourth Intl. Conf. on Genetic Algorithms, California, 1991, pp. 392–397.Google Scholar
  27. [27]
    H. S. Stone, High Performance ComputerArchitectures, Addison-Wesley, 1971.Google Scholar
  28. [28]
    B. Stott, Review of load-flow calculation methods, Proc. IEEE, 1974, pp. 916–927.Google Scholar
  29. [29]
    B. Stott and O. Alsaç, Fast decoupled load flow, IEEE Trans. Power Apparatus and Systems, PAS-93, 1974, pp. 859–869.CrossRefGoogle Scholar
  30. [30]
    S. N. Talukdar, S. S. Pyo, and R. Mehrotra, Designing algorithms and assignments for distributed processing, Tech. Rep. EPRI EL-3317 RP17643, Electric Power Research Institute, 1983.Google Scholar
  31. [31]
    S. N. Talukdar, V. C. Ramesh, and J. C. Nixon, A distributed system of control specialists for real-time operations, in Proc. Third Symp. Expert System Appl. to Power Systems, Japan, 1991.Google Scholar
  32. [32]
    S. N. Talukdar, V. C. Ramesh, R. Quadrel, and R. Christie, Multiagent organizationsfor real-time operations, in Proc. of the IEEE, 1992, pp. 765–778.Google Scholar
  33. [33]
    M. H. M. Vale, D. M. Falcão, and E. Kaszkurewicz, Electrical power network decomposition for parallel computations, in Proc. Int’l. Conf. Circuits and Systems (ISCAS), IEEE, San Diego, California, 1992.Google Scholar
  34. [34]
    A. Vanelli, Solution techniques for 0–1 indefinite quadratic programming problems with applications to decomposition, PhD thesis, University of Waterloo, 1983.Google Scholar
  35. [35]
    A. J. Wood and B. F. Wollenberg, Power Generation, Operation and Control, John and Sons, 1983.Google Scholar
  36. [36]
    F. F. Wu, Theoretical study of the convergence of the fast decoupled loadflow, IEEE Trans. Power Apparatus and Systems, PAS-96, 1977, pp. 268–275.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2002

Authors and Affiliations

  • Benjamín Barán
    • 1
  1. 1.National Computing CenterNational University of AsuncionSan LorenzoParaguay

Personalised recommendations