An integrated approach to cellular mobile communication planning using traffic data prestructured by a self-organizing feature map

The planning decision for the locations of the radio base stations in a cellular mobile communications network is the crucial point in the design phase of the network. An algorithm is presented which determines the locations of the radio base stations using a set of sensory neurons, which detect the field strength of the radio wave propagation. The sensory neurons themselves are distributed according to the traffic density, which is learned by a self-organizing feature map. Under the assumption that there exist effective methods to estimate or measure the real traffic density in the covering area, as well as to calculate the field strength at the different positions of the neurons, the presented approach provides a high-performance alternative to existing concepts