The article is somewhat confusingly worded (the actual paper it links is more clear: http://networks.cs.northwestern.edu/publications/wifm/icnp2015-flores.pdf h

The article is somewhat confusingly worded (the actual paper it links is more clear: http://networks.cs.northwestern.edu/publications/wifm/icnp2015-flores.pdf http://networks.cs.northwestern.edu/publications/wifm/icnp20...) and I initially thought it was describing something like Microsoft's White Spaces project where devices would automatically transmit over unused spectrum. In reality, the idea is that the devices all automatically select timing relative to an FM Radio Data System transmission and then can infer each other's chosen timing without actually communicating (or they "implicitly communicate" in the words of the paper's author as quoted in the article).