Packet switched communication network designs using GA

In general, the designs of mesh communication networks are complex, multiconstraint optimization problems. For a practical problem of a modest number of nodes, approximate solutions are obtained through heuristic methods. In this paper, we present a genetic algorithm (GA) solution method for the topological design of a packet switched communication network. The advantages of using GA are not only the elegance and simplicity in its procedure, but also its ability to handle continuous and discrete link capacities, linear or discrete cost structures, and different constraint models. Furthermore, its capability for designing large realistic networks is also demonstrated.