These examples in a radiocommunication context show the existing collaboration of the Computer Graphics
's and Syscom
's teams of the XLIM/SIC laboratory. In a near future, some illustrations with acoustical or wireless optical physics will be provided too.
In a RaPSor simulation, the user has to chose the numbers of reflection, transmission and diffraction.
On the example given above, the user constrainsts are a single reflection, and no transmission or diffraction. The white sphere represents the transmitter;
of course, it is possible to specify its 3d cartesian location.
Coverage area usage
RaPSor can be used for coverage area simulation: then, computation are not done for a unique receiver,
but for a set of sensors uniformely placed on a regular grid. The coverage area given above as example is produced with a central frequency of 3.5GHz,
with one reflection and one diffraction, and a 1 centimeter cover step.
In this second figure, a coverage area is depicted in 3 dimensions,
where each small sphere represents a receiver. The received power are computed in dB, whatever the used physic is.
The user can choose the used frequencies.
For instance, the figure given above represents an indoor scene with a 433Hz frequency, in a radiowave configuration.
Above is the same scene, but using a 868Hz frequency, showing the frequency modification impact.
Here, the user works at a 2.5GHz frequency, corresponding to Wifi bandwith.
Pertaining to the used physics, many sensors are proposed. For instance, in radio we can choose between: isotropic antenna, horn, dipole ...
The upper figure illustrates the horn case.
Various exploitation modes
Simulation results can be saved in XML.
In the above example are shown: the transmitter and receiver locations, and for each path, the received power, the number and description of the geometric interactions.
Another RaPSor solution is to create the impulse response from the simulation results.
We can observe above an impulse response after a rectangular filter product.
The same impulse response multiplied with a Hamming filter is shown above.
Four filters are currently proposed: rectangle, Hamming, Gaussian, and triangle.