Evolutionary Computation Techniques for Telephone Networks Traffic Supervision Based on a Qualitative Stream Propagation Model

Abstract

Evolutionary computation techniques have received a great deal of attention regarding their potential as optimization techniques for complex functions. In this paper, we consider three of them: multiple restart hill-climbing, population-based incremental learning and genetic algorithms. Their binary version and a real-coded variant of each of these techniques are experimented on a real problem: traffic supervision in telephone networks. Indeed, this task need to determine streams responsible for call losses in a network by comparing their traffic values to nominal values. However, stream traffic values are not directly available from the on-line data acquisition system and, hence, have to be computed by inverting a computational model of stream propagation in circuit-switched networks only based on the Erlang’s formula plus qualitative knowledge about the network. Then, our stream propagation model inversion has been computed thanks to the previous techniques and using several fitness measures to show how their choice can impact on the final results.