Heuristics for Evolutionary Off-line Routing in Telecommunication Networks

O -line routing in backbone telecommunications networks is a combinatorial optimization problem well-suited to iterative search methods. In this paper, a number of further heuristics applicable to the routing problem are introduced and evaluated. The presented results show that these methods signi cantly improve the search for solutions, particularly when on-line performance is considered. The use of delta-evaluation of solutions in order to reduce computation time further, is also investigated. Previously, simulated annealing has had a signi cant advantage over genetic algorithms on problems where deltaevaluation is applicable, because genetic algorithms employing standard crossover operators cannot make use of the technique. However, we introduce a specialized recombination operator which enables a genetic algorithm to exploit delta-evaluation e ectively on the routing problem. The performance of the genetic algorithm employing the new recombination operator is found to be significantly better than that of a mutation-only genetic algorithm.