Path finding based on Monte Carlo techniques compared with a full ray-tracing approach in narrow and wide bands

Efficient propagation prediction requires a tool both fast and accurate, which are two opposite features. Generally, such a tool is fast using crude approximation, leading to missing of some propagation paths. Moreover, its accuracy is usually validated by some comparisons to measurement. Then, only the statistical behavior of such a tool is available, but not its particular one for a given scene. We propose a new optimization for finding the propagation paths, based on Monte Carlo techniques. It allows a theoretical knowledge of the error due to our approximation, so that any engineer can really choose between faster or accurate calculations. Our approach is compared with a full ray tracing tool. This allows us to discuss the validity of our model, and its gain in terms of computation time. Comparisons are produced both in narrow and wide bands.