A ray launching-neural network approach for radio wave propagation analysis in complex indoor environments

A novel deterministic approach to model the radio wave propagation channels in complex indoor environments reducing computational complexity is proposed. This technique combines a neural network and a 3D ray launching algorithm in order to compute wireless channel performance in indoor scenarios. An example of applying the method for studying indoor radio wave propagation is presented and the results are compared with a very high resolution fully three dimensional ray launching simulation as the reference solution. The new method allows the use of a lower number of launched rays in the simulation scenario whereas intermediate points can be predicted using neural network. Therefore a high gain in terms of computational efficiency (approximately 80% saving in simulation time) is achieved.