Project Update - RF Propagation 


I have made some good progress towards simulating the propagation of radio waves in urban environments.  Here are some of the highlights:

Model Acquisition and Conversion 

I acquired some RF propagation measurement data collected in Munich Germany, including: a building database, collection routes, and signal strength measurement values along the routes.  The building database was not in a standard format, and so I had to write some custom code to convert the data into a standard .obj file format.  Below is a rendered image of the resulting building data.   Note the cathedral on the right-hand side of the image.  All of the buildings are just extrusions to a certain height for a given polygon footprint.  I am ignoring ground elevation changes initially, and assuming the ground is a flat plane.  Later I hope to incorporate terrain for better results.   
Nice image of munich geometry.


I obtained the RESound code used by GAMMA for sound propagation simulation.  It was not trivial to get the correct version of Visual Studio installed, and get all the settings correct so that RESound could build and run on my laptop.  I also spent some time learning how to navigate RESound's GUI, and how to use the many options, and configuration files.  In addition I familiarized myself with the audio ray tracing engine which is a part of RESound.

Munich in RESound 

I imported the Munich model into RESound, and created the necessary source, position, and other configuration files.  Several configurable values had to be tweaked in order to get RESound to work well for the urban model, including: reflection order, number of rays cast, and max attenuation distance.  The image below shows several specular reflection paths resulting from tracing a source (red) and receiver (yellow) in the same open area of the Munich model.  I traced only specular reflection path, as diffraction is currently not working in the RESound ray tracing code (expected to be fixed very soon).  

Munich propagation

The following is another screenshot from the Munich model inside of RESound, but this time the receiver is further down the street along the data collection route.  Notice that the signal must reflect at least 4 or 5 times to reach the receiver.  Without diffraction it is difficult to get good results with the receiver further away, since it becomes more and more difficult to find a specular path through the city.  Diffraction would allow the rays to diffract over the buildings.  I can now obtain a list of propagation paths for a given source and receiver, next I need to calculate the path loss due to each path.  
Munich propagation from above


I have also continued reading papers dealing with urban RF propagation.  There are several interesting methods, including dominant path prediction, which are of interest.  

What's Next

The following is a list of the next steps for this project: