Dynamic routing in homogenous ATM networks using genetic algorithms

The selection of an appropriate route for a given virtual circuit (VC) request along a set of virtual paths (VPs) satisfying the required quality of service (QOS) constraints is an important issue in asynchronous transfer mode (ATM) networks. In this paper, we propose a new routing algorithm called the Genetic Routing Algorithm (GRA) using genetic algorithms to find a sub-optimal routing solution for ATM networks. New encoding and selection mechanisms are proposed. After making suitable simplifying assumptions, simulations are carried out to implement and evaluate the proposed routing algorithm and compare it with other well-known algorithms. We found that GRA distributes the traffic load among all VPs fairly and hence minimizes both the average cell delay and the average cell loss probability.