We are talking about three different things here:
1. Ray-Tracing, a technique to simulate light rays bouncing through a scene, hugely computationally intensive, barely practical for video game rendering until recently. It creates photo-realistic images that can have DoF (trust me, I literally built my own ray tracer before)
2. Ray-Casting, a technique to solve different problems in 3D video game design where you use ray intersection to solve problems. Examples: seeing what the player is looking at, you fire a ray to see what objects get hit. this can be used with eye tracking to see what you are looking at
3. Post-processing DoF algorithms, post processing algorithms are algorithms run on the images pixels after its been rendered through raster graphics or ray tracing, examples of it that approximate DoF are the 3 algorithms in the paper thats been linked, these have nothing to do with ray tracing or casting. They are simply a fast way of calculating an approximation of DoF for real time video games.