Live data from Hacker News

What Does It Take to Track a Million Cell Phones?

thehftguy.com

31–40 of 88 posts

Re: What Does It Take to Track a Million Cell Phones?

#32

Earlier quoted context omitted.

Yes, if you're in a city you're tracked constantly by dragnet surveillance. Questions like this one aren't very useful without a threat model. Who are you trying to prevent tracking you? If it's just your phone carrier then obviously turning off your phone and removing the battery will render it inoperable. But now you don't have a phone, and your location info wasn't very useful to begin with anyway unless you were…

This goes into I Don't Know What I Don't Know Dept. (sorry Mad magazine) A patron was telling me that the way the GPS is so accurate is because it uses the phones radio... Didnt know that either. (i mentioned to him that there is one spot in MA where our google directions are off by 1/2 mile.. Same place every trip.)

It's called Assisted GPS: https://en.wikipedia.org/wiki/Assisted_GPS

It speeds up getting an accurate location, doesn't provide a more accurate location than GPS.

GPS on is generally accurate to around 5 or 6 metres. That's the technology on it's own.

Re: What Does It Take to Track a Million Cell Phones?

#33
post #20

Somebody please explain this line from the post: Radio waves travel at the speed of light 299 792 458 m/s.

Both radio waves and light are just differing frequencies of Electromagnetic Radiation.

GBNST (guilt by non-scientific thinking): Radio waves = sound = speed of sound, therefore wtf sound = light, but now I 'see the light.'

After having a second cup of coffee I did a doh! and realized conflating 'radio' with sound is non-sensical, but I wonder if I'm in the minority thinking this way. Or maybe it's just my non-tech background!

Re: What Does It Take to Track a Million Cell Phones?

#34
post #28

Earlier quoted context omitted.

> The phone companies already do this, more or less, as is shown in court cases where cell phone records are brought in as evidence You can also arrange to buy this information. I worked for a place where you could request someone's location by phone number. There were a lot of contractual obligations around us having the phone owner "allow" us to do that, but no technical ones.

Was it a phone operator specific system? Like could you get the location of any phone number or only AT&T's, for example.

I believe our provider had contracts with the big 4 in the US.

Re: What Does It Take to Track a Million Cell Phones?

#35

Earlier quoted context omitted.

> The phone companies already do this, more or less, as is shown in court cases where cell phone records are brought in as evidence You can also arrange to buy this information. I worked for a place where you could request someone's location by phone number. There were a lot of contractual obligations around us having the phone owner "allow" us to do that, but no technical ones.

What's the accuracy on that location information? Is it down to the meter/10 meter/100 meter?

It wasn't GPS accurate, but accurate enough for our usage (monitoring volunteers). I want to say it was under 10m and over 5m on average. It's been about 6 years since I've dealt with the system, so the details are a bit fuzzy.

Re: What Does It Take to Track a Million Cell Phones?

#36

If my phone is powered off can i be tracked? What if i remove the battery?

If your phone is powered off or in airplane mode it is not supposed to emit RF and thus cannot be tracked. This is a matter of trust, so if your threat model includes high end threats, the assumption that it follows the normal requirements may be invalid.

If you remove the battery, it will be unpowered and unable to emit RF and thus cannot be tracked. While it is theoretically possible to hide an auxiliary battery in your phone, that would be very hard to achieve, especially in modern thin phones. If your threat model includes highly motivated state sponsored actors, this is could be achieved.

If you put your phone in a RF-tight enclosure (e.g. metal box), the RF energy cannot get out and thus it cannot be tracked.

Re: What Does It Take to Track a Million Cell Phones?

#37
post #2

The phone companies already do this, more or less, as is shown in court cases where cell phone records are brought in as evidence A decade ago that data was a little more iffy (i.e. it was more a good estimate (typically within half a mile or less) than a true location), but with a combination of more towers (and therefore more data points), the ubiquity of smartphones (which check in more often, are doing geolocatio…

Here's one such testimony from Aaron Hernandez's murder trial, where his cell phone pinged a tower near the scene of the crime:

https://www.youtube.com/watch?v=Am8izKu5ZSU

https://en.wikipedia.org/wiki/Aaron_Hernandez#2013_murder_of...

Re: What Does It Take to Track a Million Cell Phones?

#38

This article finally answered a question I've had for a while: how they can do decent triangulation with just two towers. > "We said that a tower covers a radius around it. In practice, this is sub optimal so that’s not how it’s done. > Instead, a station is usually split in 3 independent beams of 120 degrees." So it's not the intersection of two circles anymore, it's the intersection of two arcs, which will likely o…

Sometimes it's split even more: https://www.google.ca/search?q=sector+antenna&tbm=isch

Re: What Does It Take to Track a Million Cell Phones?

#39
I've designed systems that do this on continental scales (i.e. hundreds of millions of cell phones simultaneously, in real-time). The devil is in the details and non-trivial; this is not a "an intern and 6 months" job. Mobile telemetry is not nearly as ideal in practice as assumed here and it typically takes a couple years to learn how to handle the numerous peculiar artifacts of that data that will damage the quality of a naive implementation. Reconstructing a model of the population from the cleaned data that approximates the ground truthing is surprisingly difficult and requires quite a bit of clever data science and maths.

It takes a lot of work and expertise to build a population model from mobile telemetry that approximately reflects reality. Far fewer people know how to do this well than you might assume by looking at the requirements for a naive implementation. Even most mobile carriers have limited ability.

Re: What Does It Take to Track a Million Cell Phones?

#40
this method will only work with GSM network because 1.GSM networks doesn't verify BTS 2.GSM encrypt keys are cracked and all over the internet. Users of other kind of networks should not worry about this kind of hack. Actually here in China a fake BTS a.k.a 伪基站 can be easily purchased online.
Post reply on HN