An evolutionary approach for configuring economical packet switched computer networks

The topological design of computer networks essentially consists in finding a network topology which minimizes the communication costs, taking into account some constraints such as performance and quality of service. This optimization problem is well known as difficult to solve, such that only heuristic methods are usually recommended and used. These methods are incremental in the sense that they take a starting topology as input solution and perturb it in order to produce a better solution. In this paper, we propose an evolutionary approach, based on the genetic algorithm paradigm, for solving this problem. Simulation results confirm the appropriateness and efficiency of this approach which yields solutions of very good quality for moderate size networks.

[1]  R. Boorstyn,et al.  Large-Scale Network Topological Optimization , 1977, IEEE Trans. Commun..

[2]  Lawrence. Davis,et al.  Handbook Of Genetic Algorithms , 1990 .

[3]  Leonard Kleinrock,et al.  Queueing Systems - Vol. 1: Theory , 1975 .

[4]  Amitava Dutta,et al.  Integrating Heuristic Knowledge and Optimization Models for Communications Network Design , 1993, IEEE Trans. Knowl. Data Eng..

[5]  Martin Grötschel,et al.  Integer Polyhedra Arising from Certain Network Design Problems with Connectivity Constraints , 1990, SIAM J. Discret. Math..

[6]  Samuel Pierre,et al.  A knowledge-based system with learning for computer communication network design , 1990 .

[7]  Irina Neuman,et al.  A system for priority routing and capacity assignment in packet switched networks , 1992, Ann. Oper. Res..

[8]  Bezalel Gavish,et al.  Routing in a Network with Unreliable Components , 1988, IEEE Trans. Commun..

[9]  Hideo Miyahara,et al.  Fault tolerant packet-switched network design and its sensitivity , 1991 .

[10]  Doan B. Hoang,et al.  Joint Optimization of Capacity and Flow Assignment in a Packet-Switched Communications Network , 1987, IEEE Trans. Commun..

[11]  B. Gavish.  Topological design of computer communication networks — The overall design problem , 1992 .

[12]  Mario Gerla,et al.  The design of store-and-forward (s/f) networks for computer communications , 1973 .

[13]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[14]  Samuel Pierre.  A learning-by-example method for improving performance of network topologies , 1994 .

[15]  Luigi Fratta,et al.  The flow deviation method: An approach to store-and-forward communication network design , 1973, Networks.

[16]  Samuel Pierre.  Application of artificial intelligence techniques to computer network topology design , 1993 .

[17]  D. J. Cavicchio,et al.  Adaptive search using simulated evolution , 1970 .

[18]  Lawrence Davis,et al.  Genetic Algorithms and Communication Link Speed Design: Theoretical Considerations , 1987, ICGA.

[19]  Aaron Kershenbaum,et al.  Telecommunications Network Design Algorithms , 1993 .

[20]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[21]  Gilbert Syswerda,et al.  Uniform Crossover in Genetic Algorithms , 1989, ICGA.

[22]  Samuel Pierre,et al.  An Artificial Intelligence Approach to Improving Computer Communications Network Topologies , 1990 .

[23]  Stephen F. Smith,et al.  A learning system based on genetic adaptive algorithms , 1980 .

[24]  M. Schwartz,et al.  Routing Techniques Used in Computer Communication Networks , 1980, IEEE Trans. Commun..

[25]  Mario Gerla,et al.  On the Topological Design of Distributed Computer Networks , 1977, IEEE Trans. Commun..