Mozilla Is Mapping Cell Towers and WiFi Access Points 113
First time accepted submitter neiras writes "Mozilla is building a map of publicly-observable cell tower and WiFi access points to compete with proprietary geolocation services like Google's. Coverage is a bit thin so far but is improving rapidly. Anyone with an Android phone can help by downloading the MozStumbler app and letting it run while walking or driving around. The application is also available on the F-Droid market." "Thin" is relative; it's quite a few data points since we first mentioned the pilot program a few months ago.
Re:Privacy (Score:4, Informative)
Re: (Score:2)
Can you really help intercepting it in some fashion if you're mapping networks? What you can help is storing it afterwards. It should have been discarded in microseconds, not months.
Re: (Score:3)
Re: (Score:3)
You don't need to 'connect' to them but IIRC there is some benefit to looking at the traffic beyond mere broadcasts. IE if you can see device X sending traffic to Y you can begin to imply the position of Y even if you can't see it that device yourself because it's too far away from you.
A <------ X <-------> Y
Moz might not be doing that and perhaps it isn't a
Re: (Score:3, Informative)
Also to improve GeoIP: If they connect to the WiFi, know the IP it uses to get to Google servers. Then they can provide the most probable location data back when some device on that IP asks for location information.
Re: (Score:3)
That is going to be of limited use though. There is a good but less than 100% chance that the AP is stationary, There is a good bit less chance the thing it is talking to are stationary.
There is more safety in sticking with beacon packets as well. It's really hard to claim that a beacon was meant to be private.
Re: (Score:2)
Re: (Score:2)
That could be somewhat useful, but would be a much lower precision so would need to be noted a such.
Re: (Score:2)
Re: (Score:2)
RTFA or at least RTFS
Re: (Score:2)
If one needs to interface to the internet, would knowing where an access point be useful? One either has access or doesn't. If you don't, then move till you do.
Now if Mozilla wants to get into the Cell Tower, WiFi business, then their actions make sense. All Cell Towers are registerd with the government. WiFi's are registered with their various ISP providers. So Mozilla can't go to those folks that supply this service and ask? Why? Is it some big scarey secret
Re: (Score:2)
Perhaps once on the mozilla map page (linked in TFS) you might have selected Project overview [mozilla.org] to see what they had in mind if it wasn't yet clear.
The telecomms industry isn't well known for being at all helpful even when being paid.
You obviously also didn't look at my UID :-)
Re: (Score:2)
Re: (Score:1)
Indeed, your device probably does it already, to present you a list of available networks.
Re:Privacy (Score:5, Informative)
Completely. This just looks for the 'announcement' packets from access points. It doesn't care or do anything about the data packets.
You are intercepting data just as much as your phone does when you go to the wifi page and it shows the list of access points near by.
Google was accidentally storing all the raw data for debug purposes (which got left turned on).
Re: (Score:2, Interesting)
Google was storing all the raw data.
Fixed that for you.
Accidentally? (Score:2, Insightful)
Why the Google ad broker propaganda? One doesn't accidentally store such an amount of data for such an amount of time. End of. Don't echo their propaganda like some idiot fanboy.
Re: (Score:2)
One doesn't accidentally store such an amount of data for such an amount of time.
Umm, Google kept the data because they recognized that the logging was going to provoke regulatory investigations, and deleting it could be construed as destruction of evidence.
Re: (Score:1)
>What's the point of this?
RTFS, geolocation.
>Didn't the FTC fine Google for this kind of activity?
No, they got fined for a different aspect.
Re: (Score:1)
I'd say this is especially an issue with so little data currently.
I can currently narrow out individuals in my neighbourhood, the paths they took from their houses (which I can narrow down to about one of 6 addresses in two caes i'm looking at) to the local shopping center, to another persons house (friend, relative - unsure but it'd help identify the individual), and back to their house in this case.
If I got the complete data with timestamps, I could easily and completely filter out individual trips withou
Re: (Score:2)
You are right, of course, it merely follows people, it says nothing of signal paths, and can't distinguish no-signal areas from un-visited areas.
And showing a map when there are so few participants is pretty silly.
For cell reception, this is useless.
For wifi mapping, this is redundant [wigle.net].
WiGLE EULA (Score:2)
You are right, of course, it merely follows people, it says nothing of signal paths, and can't distinguish no-signal areas from un-visited areas.
If trilateration signal in a given area is marginal, the data collection should mark the area as marginal. If marginal area surrounds unvisited area, one can be fairly confident that the border is between a visited area and a no-signal area.
For wifi mapping, this is redundant
Not if Mozilla plans to make the data available to the public under terms more permissive than the WiGLE EULA [wigle.net]. It could be an example of what Google's Greg Stein called "license pressure" [slashdot.org].
Re: (Score:3)
I've been doing this for a few months. I can see the trip I took over New Years week down the California coast and then across to Death Valley and up to Lake Tahoe. I don't care that people can see my track (I spent some time in Los Osos and Morrow Bay and you can clearly see my routes in that area).
I assume that the NSA also has my route from tracking my cell phone.
Re: (Score:2)
and if I could figure out the frequency of data transmission
LOL your not trying, Google: frequency finder
But then they could catch on and use a cell phone jammer http://hackedgadgets.com/2007/12/10/cell-phone-jamming/ [hackedgadgets.com]
I was in traffic court this week (normal part of driving for me) and noticed they had a jammer in use (keeps cell phones from ringing during court).
.
Re: (Score:1)
The individuals in your neighbourhood walk around with access points? That must be a very unusual neighbourhood. Most people who have an access point just put it somewhere stationary.
Re: (Score:1)
What's the point of this?
So they can sell your location to advertisers?
Re: (Score:2)
What's the point of this?
So they can sell your location to advertisers?
I know for a fact that Mozilla doesn't want to do that, it's stated in the project goals. The point is that it's eays to find out location without having to turn onf/have access to a GPS.
Re: (Score:2, Interesting)
Presumably they do it so that they can support the HTML5 Geolocation API. And, FWIW, HTML5 Geolocation is opt-in on every request, at least in the context of general web browsing.
I've found that HTML5 geolocation is more accurate using Wifi than cell towers or even GPS. Cell towers don't give very accurate results because cellphones and tablets don't actually triangulate your position like they might do with GPS. GPS sucks because people are inside most time, and also the GPS receiver chipsets in cellphones
Re: (Score:1)
Yelp. Google Maps. Foursquare. Whatever. All written in HTML5.
IE, Chrome, Safari, and Opera all already support Wifi geolocation. Firefox does, too, actually, but Mozilla probably doesn't like the price they're paying their third-party vendor.
Re:Privacy (Score:4, Informative)
Geolocation for what though?
Say I go to google maps, or mapquest, or openstreetmap, or whatever. That web page will ask the browser to do geolocation, the browser will throw up a dialog something like this:
Website foo-maps.bar wants to know your location.
{Scare-text explaining why I might not want to click OK}
Send location data?
[Yes] [No]
If I click yes, the browser sends my location to the mapping site. Now I'm looking at a map of where I am, I can search for businesses nearby, etc.. Or, if I don't want a map of my current location, I could just click no, then I'd have to type in an address or search query to find the map I do want.
Or some ad server wants to show me banner ads for nearby stores, so it asks the browser to geolocation; the browser will throw up the same dialog, I'll click no, and the ad server doesn't get my location.
Or any other web site that might want to know my location for any reason, same story: The browser pops up a dialog, I click yes or no, and the site gets or doesn't get my current location.
Re: (Score:2)
ans you get stuck looking for ok button !
Re: (Score:1)
Re:Privacy (Score:5, Informative)
Actually, it's to... provide an alternative locating service to GPS.
Both Apple and Google maintain a list of WiFi MAC addresses and GPS locations. In areas where there's no GPS, or GPS is extremely weak, using cell tower and MAC addresses can provide alternative location services. Or for devices without GPS hardware, it can provide location services still. E.g., if you tether a WiFi-only iPad to an iPhone, it can get your location quite accurately using the database.
Apple bought a company that maintains the database, Google built theirs up using streetview. Mozilla is probably trying to create an open-source version.
And it's that database that lead to the whole "tracking" scandal of iOS 4 - because whenever you requested a location Apple sends you a database containing locations near you as well so you can do mapping without continually asking Apple where it is. That database cache was what people said "Apple is tracking them!" Of course, it wasn't, but knowing what areas the cache covers helps in knowing where you might be. In densely populated areas with a lot of APs, Apple would send you a very narrow list that can be quite accurate to your track. In areas with more sporadic coverage, you get a bigger footprint because there's less data per square mile (Apple probably sends you a fixed number of APs to locate oneself, rather than send you all the APs within a certain radius).
So in the city, you can get down to street-level tracking. In the suburbs, well, the cache is probably only good for pinpointing to a few blocks.
Re:Privacy (Score:4, Insightful)
Also, GPS kills batteries. A quick network lookup (or even local, since you could cache the local area's data and request new data only when you move enough) is cheap on the battery.
Re: (Score:3)
If you tether a Wifi only iPad to an iPhone, you can use the GPS on the iPhone to find out where you are...
Re: (Score:2)
Re: (Score:2)
Google-ish issue? (Score:1)
Hopefully they run into the same issue that Google did.
Re: (Score:3, Insightful)
Re: (Score:2)
Hopefully they run into the same issue that Google did.
Large location based advertising revenue? A global world road map with radio-location markers that exceeds many commercial cartographers' efforts? Which issue were you thinking of?
Re: (Score:2)
Maybe the one where they've got so much money that they can drop $3.2bn on a thermostat company and count it as an operating expense?
It was already there. (Score:1)
openBmap, anyone?
Maps roads, Not Coverage (Score:5, Interesting)
Its a frequent problem with these phone based mapping programs, that the coverage area they map is way too small, especially when they are mapping cell towers. They usually assume a reception circle about the width of a road. So they end up mapping roads, and frequently apply magical thinking to show no coverage areas simply because nobody walked there running their app.
They will show coverage on all sides of an open field, but unless someone walks a zigzags path thru that field they will simply assume there is no coverage there. I prefer carrier maps. Even guesswork by real radio engineers is better than spotwork by silly apps.
These mapping programs, when mapping cellular service would be better off mapping HOLES (no coverage areas) of each type (2g, 3G, LTE, CDMA, etc). The task would be smaller, and the data presentation far more useful. They would just log GPS position where there was no signal and send that when they again found a signal. Presentation would show service available until you actually had some measurements that said it wasn't.
That way at least the farmer or hunter working off road would have a more reliable idea of where there is likely cell service, and everybody would have a better idea of where they are unlikely to service.
Assuming it is all quiet in the forest when trees fall simply because you weren't standing there to hear it is a interesting philosophical exercise but a pretty stupid way to run a mapping service.
Re: (Score:3)
Mapping holes might be a smaller task in urban areas, but I assure you that's not the case in much of America. The two methods could easily be combined (map holes in urban/suburban areas, map coverage in rural areas) to make this an easier task, though.
Re: (Score:2)
Agreed, I have no problem with using both approaches.
But they should at least buy a real radio engineer a cup of coffee and find a reasonable estimation of the radius or reception around any given location when the device is measuring a given dBm. Assuming the signal falls to zero at the edge of the roadside is silly.
Re: (Score:1)
They would just log GPS position where there was no signal and send that when they again found a signal.
mod parent up... this would be extremely useful, including to those looking at possibly changing carriers
also, if you know the location of towers fairly accurately, you only need one data point to determine the reception radius all around the tower for the specific phone/device you are using
a possible complexity might be differences in reception on various devices (including possibility that "you're holding it wrong") but results could also be affected by bridges/tunnels, topography, background interference
Re: (Score:3)
also, if you know the location of towers fairly accurately, you only need one data point to determine the reception radius all around the tower for the specific phone/device you are using
Exactly.
The phone knows what tower it was connected to.
The phone knows its current signal/noise ratio.
The phone knows how much power it needs to use to be heard by the tower.
And the phone know where it is, rather precisely if GPS is on.
If you are measuring -75dBm where you are standing, its reasonable to assume a far bigger circle of reception than if you are seeing -101dBm.
In neither case is there a reason to assume reception disappears at the ditch beside the road you are walking.
This whole thing appears
It's cause you're a douche (Score:3, Insightful)
Re: (Score:1)
i see what you mean, and maybe i wasn't the target of icebike's flame but i wasn't offended since i know most of what he said (in this case) is true. i just took the last paragraph with a grain of salt (this is slashdot after all).
technically if the point 6 bales of hay into the field that he mentioned was shielded behind an iron clad building (like a hay or machinery shed) it could make a difference :-)
Re:Maps roads, Not Coverage (Score:4)
I'm not sure what your point is. This isn't supposed to be a map of cell phone coverage. It's a map showing all the data points in their database. The goal of this project is to let people identify their location based on the visible networks, not to tell them what kind of network coverage they'll have in any location.
Re: (Score:2, Informative)
Follow the first link in the story. The biggest text on the page says COVERAGE MAP and when you follow the other links
it is clear that their intent was a coverage map, not a data-point map where Joe Sixpack happened to see a Cell Tower.
Re:Maps roads, Not Coverage (Score:5, Informative)
I did follow the link. You're misinterpreting it. This is a data coverage map, that is, a map of how much data they have in different places. It has nothing to do with cell phone coverage.
Project is for GeoLocation NOT Cell Coverage!! (Score:5, Informative)
This will allow the look up of rough position information without turning on the GPS using an OPEN DATABASE. The same thing that a few PROPRIETARY databases do currently.
Given this goal, road coverage is good enough.
Re: (Score:1)
What f'ing "guesswork"? They know where the tower is, and draw a circle around it. Done. No engineer; just a database (that anyone can build from FCC data, btw) and simple program (I did it with a bash shell script.) Using Google Earth elevation data would make it a little more accurate, but that's a lot more programming.
Re: (Score:2)
If they could get that out of the FCC database, why put an app on a phone and log this.?
After all, if you look at their map, they are simply showing where people were standing (driving) at the time their phone reported, and no tower locations are shown. Look here, https://location.services.mozilla.com/map#15/47.3771/8.5373 [mozilla.com] maximum zoom into Zurich. You have streets mapped, but no tower data at all. They are replicating street maps, not tower or wifi maps.
Re: (Score:1)
You'd have to ask them; I guess they wanted in on the whole wardriving craze -- over a decade late... The FCC only covers cell sites in the USA. (nor does it say which towers are actually *on*)
Re: (Score:1)
i thought android+google maps already does this... i think they call it "coarse-location" (due to it not being as accurate as "fine-location provided by gps)
Re: (Score:1)
http://android.stackexchange.com/questions/1648/how-does-android-get-the-coarse-location [stackexchange.com]
Re: (Score:3)
i thought android+google maps already does this... i think they call it "coarse-location" (due to it not being as accurate as "fine-location provided by gps)
They do. And if you RTFS, Mozilla is also doing it "... to compete with proprietary geolocation services like Google's."
Re: (Score:2)
to compete with proprietary geolocation services like Google's
how is mozilla going to "compete" with something that's already second to none, comes pre-installed on android handsets, and is free to use with no intrusive ads?
providing an alternative is fair enough (like the choice of linux distros) but if you understand the technology behind existing coarse location services already built into android handsets, what additional value is the mozilla app realistically likely to add?
maybe their slogan could be "but at least we're not google" because that's pretty much the
For Firefox OS and Android devices w/o Google Play (Score:4, Insightful)
how is mozilla going to "compete" with something that's already second to none, comes pre-installed on android handsets, and is free to use with no intrusive ads?
By making it available on Firefox OS handsets and on those Android devices that don't ship with Google Play Services, such as Kindle Fire series, several devices popular in China, and phones with CyanogenMod system software installed that don't have the Gapps.
Re: (Score:1)
hmm... fair enough :-)
Re: (Score:2)
And while Google are obviously willing to license usage of this to some extent (e.g., Presto-Opera's geolocation used the Google coarse location API), relying on licensing something from a third party (and one whom is frequently a competitor, given the number of markets Google are now in) is risky, especially given Google has fairly aggressively deprecated APIs before, at times without replacement.
_nomap (Score:1)
Re: (Score:3)
I think one purpose of this is to help refine GPS position. If you know the locations of SSIDs then you can get a better location. No need to access the WiFi.
Re: (Score:1)
An open database for position estimation *without* GPS.
If you have a working GPS signal (well, technically three+ signals from the birds), you already know where you are.
Re: (Score:2)
GPS can take a while to get a fix if it has been some time or distance since last turned on.
Having an accurate location from a known WiFi spot can help the GPS get its bearings faster.
Re: (Score:1)
Re: (Score:2)
t would seem that its essentially useless as soon as a certain percentage of SSIDs change.
Then you map it again. It's not that hard especially if it works ok for ~2 years, and you manage to get an installed base during that time. Every time a phone says "i can see X Y Z", and Z isn't in the DB you will now that Z is near X Y.
Re: (Score:1)
When you record your known position as well as the BSSIDs and cell towers, and the signal strengths that you are seeing, when someone else in the future queries the database with "I'm seeing these BSSIDs and those cell towers, where the hell am I?" the database can give an answer. Kind of like GPS, but without the need to see the sky and with less battery drain on your device.
Partner up (Score:3)
http://sensorly.com/ [sensorly.com]
Has already done much of what this project is wanting to accomplish
Re: (Score:1)
just point them to google maps, which takes advantage of android coarse location services already
obligatory (Score:1)
http://xkcd.com/407/ [xkcd.com]
In Soviet Russia, Mozilla maps YOU! (Score:1)
There are political activists in Russia and worldwide who believe that mesh networks are usable as a backup communication medium during Internet blackout caused by political instability. The databases of WiFi geomapping just help the opponents to disrupt the backup communications.
OpenSignalMaps (Score:3)
How is this any different from the OpenSignalMaps project?
Re: (Score:2)
app for n9 ? (Score:2)
so, is there an app for n9 to contribute ? :)
Re: (Score:2)
...although i might wait anyway until mozilla resolves concerns raised at http://pavelmachek.livejournal.com/120952.html [livejournal.com] (mentioned in http://tech.slashdot.org/comments.pl?sid=4681671&cid=45995645 [slashdot.org])
Re: (Score:1)
Re: (Score:2)
interesting, although it might be over my head to get that working :) ;)
btw, there's a typo in the readme - dfrom
Illegal in some countries (Score:2)
Re: (Score:1)
Google agreed with the Dutch Data Protection Authority to let people opt out by appending _nomap to their wifi ssid. Mozilla simply has to support that optout to avoid trouble.
Re: (Score:2)
paranoid people shouldn't use Wifi. (Score:2)
Considering how easy it is to map APs I don't see why they should have to care about _nomap.. I don't see why you would want opt out.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
the easiest way to handle google bots is to display a fuck off message to any user agent containing a blacklisted string
there's websites that list all the different known user agents that you can use as a guide
the fuck off message that i use in my websites is:
"The user agent with which you have accessed this website is banned from accessing content as it is suspected of attempting to steal copyright information without the copyright holder's permission."
We still need POA (Score:2)
Google owns all the location addresses in the world. Yelp holds a much smaller subset. I care less for WiFi location capability (everyone has a GPS) than I do for being able to look something up.
Re: (Score:1)
Re: (Score:1)