Genetic Algorithm-based approach to allocation of distributed objects using graph models

  • Authors:

  • Seunghoon Choi;Jaewon Oh;Chisu Wu

  • Affiliations:

  • School of Computer Science, Duksung Women's University, 419, Ssangmun-Dong, Dobong-Gu, Seoul, Korea;SE Lab., School of Computer Science and Engineering, Seoul National University, San 56-1, Shilim-Dong, KwanAk-Gu, Seoul, Korea. Tel.: +82 2 874 4165/ Fax: +82 2 887 8991/ E-mail: jwoh@selab.snu.ac ...;SE Lab., School of Computer Science and Engineering, Seoul National University, San 56-1, Shilim-Dong, KwanAk-Gu, Seoul, Korea. Tel.: +82 2 874 4165/ Fax: +82 2 887 8991/ E-mail: jwoh@selab.snu.ac ...

  • Venue:
  • Integrated Computer-Aided Engineering
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, the partitioning/allocation model and algorithm for mapping object-oriented (OO) applications into the heterogeneous distributed environments are proposed. Our model applies the graph-theoretic approach, dealing with a lot of characteristics of OO paradigm. In addition, individual metrics for communication cost, concurrency and load balance are defined. Our allocation algorithm is based on the Niched Pareto Genetic Algorithm (NPGA). The reason for using this technique is that a partitioning/allocation problem is multiobjective problem with non-commensurable (measured in different unit) objectives, and NPGA was proved to be effective in this kind of problem. We validated our model and algorithm by experimenting on the three typical OO systems in CORBA-based distributed environments. The main advantage of this approach is that the algorithm produces not a single solution, but a family of solutions known as the Pareto-optimal set, out of which the developer can select an optimal solution appropriate for his or her environmental conditions.