An ANN approach to spare capacity planning

An artificial neural network (ANN) model using backpropagation learning to perform telecommunication network spare capacity planning while minimizing total network cost is presented. This system, called spare capacity planner (SCP), is successfully used to perform capacity planning for high-speed digital transmission networks with digital cross-connect (DACS) equipment. The use of ANN architecture to represent and compute constrained network parameters, such as point-to-point intertrunk traffic, point-to-point local traffic, point-to-point intertrunk spare capacity, individual node basic equipment cost, individual node spare equipment cost, cost of network expansion, and physical link cost, is described.<>