WiFi probes are indeed a huge information/privacy leak. I'm guessing a lot of this has to do with the fact that WiFi was not designed from the beginning with security in mind.
Simply using something like Kismet[0] yields a lot of information, even before any rigorous analysis is done. You'll see some probe requests with huge lists of SSIDs. Some of those SSIDs are comprised of an address (presumably a home address), others are obviously office/work SSIDs and still others are public ones. (Starbucks, etc.) From this you can infer a device's movement and thus likely a person's. This is all from a superficial analysis.
Others have done much more research into large-scale collection and analysis of WiFi probes. The information you can collect is immense. See Snoopy[1] and this research paper entitled, "Signals from the Crowd: Uncovering Social Relationships through Smartphone Probes". [2]
Some of these use a SSID-to-geolocation database to assign a physical location to specific SSIDs. WiGLE is an example.[3] Google, Apple, et al. maintain their own databases that are likely more accurate, used to provide geolocation services to mobile users.
Simply using something like Kismet[0] yields a lot of information, even before any rigorous analysis is done. You'll see some probe requests with huge lists of SSIDs. Some of those SSIDs are comprised of an address (presumably a home address), others are obviously office/work SSIDs and still others are public ones. (Starbucks, etc.) From this you can infer a device's movement and thus likely a person's. This is all from a superficial analysis.
Others have done much more research into large-scale collection and analysis of WiFi probes. The information you can collect is immense. See Snoopy[1] and this research paper entitled, "Signals from the Crowd: Uncovering Social Relationships through Smartphone Probes". [2]
Some of these use a SSID-to-geolocation database to assign a physical location to specific SSIDs. WiGLE is an example.[3] Google, Apple, et al. maintain their own databases that are likely more accurate, used to provide geolocation services to mobile users.
0. https://www.kismetwireless.net/
1. http://www.sensepost.com/blog/7557.html
2. http://conferences.sigcomm.org/imc/2013/papers/imc148-barber...
3. https://wigle.net/