Modeling Wireless Links

Well, since I can't sleep tonight (too much bad news in a day) I decided to read and write about this paper.

The content is quite simple, it notes that modeling wireless links is not always intuitive, leading to poor simulations of the technology. It then proceeds to give arguments about the appropriate ways to model each particular type of wireless link property.

I found it interesting that network simulators suck so badly. They seem like simple things to model, but perhaps this was before the advent of virtualization at the scale we see today. Has this improved since the publication of this work?

It's also neat that there can be important work on simply simulating networks. Click is an example of a router infrastructure that was designed for simulation as well. Perhaps simulating networks in areas where there are not many sources of reliable data would be a valid area for work.