Yes, you could use OptiX for this kind of radio simulation easily. There are other OptiX users doing simulations very similar to what you describe. I think you could demonstrate the feasibility of this by using or making minor modifications to the samples we provide with the OptiX SDK.
By the way, you might know this already, but just in case it helps you get a head start – if your receiver is large (like a sphere surrounding the scene) then tracing rays starting at the source can be relatively efficient. But if the receiver is small then the probability of rays hitting the receiver can be very low, and a lot of work might be wasted. There is a lot of research into this problem in the ray tracing community, so it’s worth studying. We even have some examples in the OptiX SDK: the optixPathTracer sample might be a reasonable place to start. The idea is to trace rays backwards starting from the receiver, not the source. When you hit any surfaces with a ray cast from the receiver, connect a reflecting ray to the source and compute the amount of signal received using a reflection function. (It’s mathematically the same function you would use to compute a reflected ray’s direction if you start from the source and reflect rays off of any surfaces you hit.) For example, imagine you’re rendering an image, then the source is the light, and the receiver is the camera. In rendering we usually start rays from the camera and work backwards to the light, because it’s much more efficient.
–
David.