A thesis submitted in partial fulfillment of the requirement for the degree of Master of Engineering, School of Advanced Technologies

Thesis (M.Eng.) - Asian Institute of Technology, 1998

The optimum network planning methods are necessary when designing optical networks so as to minimize the overall network cost while providing an acceptable performance levels of the network. The problem is of relevance as because in recent years, there has been a major stride in the development of optical networks, which provide both switching and transmission within the optical domain and also offers an extremely large bandwidth. This study considers together the network design considerations of maximum propagation delay, traffic capacity requirements between the nodes, overall network reliability and network survivability for the optimization of an Optical Network. All these network design considerations have been integrated to a minimal set of constraints for optimization. The Genetic Algorithm (GA) is used as the optim1zation tool to obtain the minimal cost physical topology design. The cost of the network is considered as a function of the total length of the fiber to be used and the total node degree of the network. The algorithm has been tested on a 5-node and a 10-node network and also compared with some previously obtained topology. The GA has also been tested for adding extra nodes to an existing optimized network and also in the case of the increase in traffic in some nodes of an optimized network. The results show that the genetic algorithm is an effective algorithm for such problems, and possibly many other topology optimization problem and the integrated problem formulation used in this study works well with the genetic algorithm.