Planning of the infrastructure of networks FWA with genetic algorithms

by Domingos, Aline Paulino

New transmission technologies give more flexibility and efficiency to the Access network, which is supposed to deal with the continuous growth of demand for telecommunication services. Operator companies need to innovate their networks in order to provide broadband services. Therefore, the network planning is essential to achieve the best performance under minimum cost. An alternative of evolution that seems to be interesting is the provision of services through fixed wireless networks. These networks, generically known as FWA (Fixed Wireless Access), can reduce the infrastructure access cost. In this paper, we present a new approach to the planning of FWA systems. First, we develop an optimization model with binary decision variables. The mathematical formulation to the model is also stated. Because solving mixed integer problems through exact methods is a complex task, we adopt a strategy based on genetic algorithms (GA). Thus, we show how to specify the main components of a GA: chromosome, solution coding, fitness assessment, population structure, and genetic operators (crossover and mutation process), in order to specialize the technique to the FWA planning problem. Besides, chromosome feasibility is ensured. An application of the proposed method is also reported