Forgot your password?
typodupeerror
Cellphones Communications Hardware

How Mobile Phones Work Behind the Scenes 220

Posted by timothy
from the they-all-suck-to-different-degrees dept.
adamengst writes "We seldom think about how our mobile phones actually work, but in this TidBITS article, Rich Mogull pulls back the covers and peels away the jargon to explain why text messages work when voice calls are dropped, why your battery lasts longer in some places than in others, why you're not allowed to use phones on airplanes, why you can be notified of a voicemail message when your phone never rang, and more."
This discussion has been archived. No new comments can be posted.

How Mobile Phones Work Behind the Scenes

Comments Filter:
  • by Gizzmonic (412910) on Tuesday October 07, 2008 @01:25PM (#25288875) Homepage Journal

    Should I try it from my mobile phone?

    • by old7 (564621) on Tuesday October 07, 2008 @01:27PM (#25288909)
      Their web server must be a cell phone.
      • http://www.tidbits.com/about/in-use.html [tidbits.com] Emperor The machine emperor.tidbits.com, also known as www.tidbits.com and just tidbits.com, is our main server. It does basically everything for us now.

        Dual 1.33 GHz Xserve G4 - [Our server, sic]Emperor runs on a normal dual 1.33 GHz Xserve G4 (2 GB of RAM). Emperor is still running Mac OS X Server 10.2.8, which came with it and handles the load just fine, so we haven't had any reason to upgrade.

        Web Crossing - The server software that powers all of our Inter
    • by electrictroy (912290) on Tuesday October 07, 2008 @01:33PM (#25288997)

      >>>Or why a text message can get through when a call can't?

      This is no great mystery. A test message can just sit in a buffer until your phone is within broadcast distance, and then it's sent. But a call has to be done in realtime; if reception is poor the caller gets a busy signal (and then send a text instead).

      • by nwf (25607) on Tuesday October 07, 2008 @01:38PM (#25289109)

        This is no great mystery. A test message can just sit in a buffer until your phone is within broadcast distance, and then it's sent. But a call has to be done in realtime; if reception is poor the caller gets a busy signal (and then send a text instead).

        And they require much less bandwidth and don't tie up a phone line out of the cell tower. Just data, which can go over a shared data line asynchronously.

        • by Volante3192 (953645) on Tuesday October 07, 2008 @01:54PM (#25289359)

          And yet still cost more than an actual call...

          • Re: (Score:2, Informative)

            by Billhead (842510)
            Not at all.
            Normally, you pay for a voice plan, and if you go over you get charged a ridiculous amount per minute.
            There is typically either no text messages included in that plan, or something like 200.
            I use Sprint, and here is their prices:
            Unlimited everything (the only way to get unlimited voice) - $100 per month.
            Adding unlimited text messages to a normal plan - $20 per month.
            If you don't have an unlimited voice plan, you get charged around $.40-$.45 per minute over, twice as much as the $.20 for a t
            • Unlimited everything (the only way to get unlimited voice) - $100 per month.

              If only. The big carriers are unlimited VOICE for $100/mo. AT&T, e.g., adds on another $35/mo for unlimited messages/data.

              Sprint does give you unlimited everything for $99/mo. Unfortunately, whether it's my phone or their network (or both), it sucks royally. As soon as one of the other carriers gets closer to comparable, I'm out of there ($175 term fee is only $50 than my one-month bill, after taxes and fees).

            • If you don't have an unlimited voice plan, you get charged around $.40-$.45 per minute over, twice as much as the $.20 for a text message.

              With those numbers voice costs more then text only if you assume one text message is equivalent to 30 seconds worth of conversation. In practice I doubt that's the case.

          • Re: (Score:3, Informative)

            by TooMuchToDo (882796)

            The ratio of signaling channels to voice channels is something around 1:21, hence the signaling channel is a scare resource compared to the voice channels (and therefor more expensive than voice calls).

            • Re: (Score:3, Insightful)

              by ceoyoyo (59147)

              Except that a text messages takes FAR less than 1/21 of the data that a similarly priced voice call does.

        • by mc900ftjesus (671151) on Tuesday October 07, 2008 @02:35PM (#25289969)

          Text (SMS) are sent over paging channels, not data channels. This is why they're still 160 characters. Yes, it's data but it's send in messaging protocols used for voice signaling. They can still get through if there are no voice channel available since they never need to setup a whole call.

          Telecom is old, don't assume things work the way they seem to as lots of legacy protocols are still in use.

        • by Piranhaa (672441) on Tuesday October 07, 2008 @02:51PM (#25290229)

          You forgot the best part about texting. Assuming it's not at night, you can do it more inconspicuously while driving!

        • by afidel (530433)
          Actually, pure GSM SMS is carried in the status message that the phone and tower trade whenever they talk which is why it is so limited in length. Today most phones will attempt to use the data service if available and fallback to the old method only if the data connection doesn't work.
    • Working link (Score:3, Interesting)

      by againjj (1132651)

      http://db.tidbits.com/article/9796?print_version=1 [tidbits.com]

      Odd that the print version on the same site works.

  • by iamdrscience (541136) <michaelmtripp@g m a i l . com> on Tuesday October 07, 2008 @01:30PM (#25288963) Homepage
    The real answers:

    why text messages work when voice calls are dropped

    Text messages are magic.

    why your battery lasts longer in some places than in others

    Some places are magic.

    why you're not allowed to use phones on airplanes

    Pilots are afraid of magic.

    why you can be notified of a voicemail message when your phone never rang

    Voicemails are magic.

  • by bendodge (998616) <[bendodge] [at] [bsgprogrammers.com]> on Tuesday October 07, 2008 @01:37PM (#25289101) Homepage Journal

    And here we see illustrated why a reading the article isn't always a good thing. This summary is obviously designed to drive people to the site hosting this article (and lots of ads I'm sure), but by forcing people to read the article you've taken down your site and most of us will now leave this page. Nice.

    On a side note, what we do have in the way of a summary suggests that there's very little for us to learn here.
    1. Text messages work when voice calls are dropped for the same reason Morse can get through when SSB voice can't.
    2. Your battery lasts longer in some places than in others because the phone automatically adjusts its transmit strength based on the distance from the tower.
    3. You're not allowed to use phones on airplanes because of paranoid ignoramuses and the insightful people who realize how bad it could get when people in a flying bomb know what's going on (and how annoying cell phones are).
    4. You can be notified of a voicemail message when your phone never rang because the network was too busy to initiate the connection, your phone was on vibrate or it didn't have a connection at the moment.

    There. Now you can get on with your day.

    • by mmontour (2208) <mail@mmontour.net> on Tuesday October 07, 2008 @01:47PM (#25289259)

      3. You're not allowed to use phones on airplanes because of paranoid ignoramuses and the insightful people who realize how bad it could get when people in a flying bomb know what's going on (and how annoying cell phones are).

      Or, just possibly it's because:
      1. GSM phones are known to emit strong pulses of RF that interfere with nearby electronics (audio amplifiers, televisions, speakerphones, etc).
      2. Airplanes contain quite a few important electronic systems for navigation, communication, flight control, etc.
      3. Considering the number of passengers who are carried by airplanes each year, even something with a one-in-a-million chance of causing a problem would be a very bad thing.

      • by Free the Cowards (1280296) on Tuesday October 07, 2008 @03:31PM (#25290763)

        Considering the number of passengers who are carried by airplanes each year, even something with a one-in-a-million chance of causing a problem would be a very bad thing.

        This is blatantly false. Airliners are chock full of things with much higher odds of failure than one in a million. Airliners achieve their extremely good safety record through redundancy and robustness, not through avoiding failure at all costs. Airliners have things fail all the time, it's just that the vast majority of the time the inherent redundancy and robustness of the aircraft make it such a minor event that the passengers don't even know it happened.

      • by tsa (15680)

        Another thing is that in the case of a crash, you don't want all those mobile phones flying through the plane and injuring people. That's why you are not allowed to use electronic equipment during takeoff and landing.

        • Another thing is that in the case of a crash, you don't want all those mobile phones flying through the plane and injuring people.

          You must be kidding me...

          I think it'd be somewhat fatuous to worry about a flying game boy than, you know, the whole "deceleration trauma" thing as the ground decides to shove itself up my posterior at terminal velocity.

          • by tsa (15680)

            So you'd rather survive the crash with a bleeding headwound from a gameboy and no further injuries than just be able to normally use the emergency exit?

            • Uh... if I survive a plane crash at all, I'm not about to be picky about minutiae like bleeding wounds, broken limbs, or a ruptured spleen.

    • by bws111 (1216812) on Tuesday October 07, 2008 @01:49PM (#25289289)
      Well, at least your number 3 is wrong. Cell phones are not allowed on planes because a few hundred phones simultaneously hopping from tower to tower at several hundred MPH wreaks havoc on the phone system. It is an FCC rule, not an FAA rule.
      • Yet I bet every day, 1,000's of phones are left on in the air unintentially, without consequence.
        • Re: (Score:2, Insightful)

          by bws111 (1216812)
          That may well be, but even 1000s of phones spread out over a day is not much compared to 100s of phones on every plane.
      • by ptbarnett (159784)

        It is an FCC rule, not an FAA rule.

        Actually, it's both:

        The FCC rule says you can't use a cell phone while airborne.

        The FAA rule says that the operator must prohibit the use of electronic devices in flight, unless the OPERATOR is sure that they will not interfere with the aircraft systems.

    • Phones on airplanes (Score:5, Informative)

      by dj245 (732906) on Tuesday October 07, 2008 @01:53PM (#25289343) Homepage
      3. You're not allowed to use phones on airplanes because of paranoid ignoramuses and the insightful people who realize how bad it could get when people in a flying bomb know what's going on (and how annoying cell phones are).

      This only half the story. There are a couple technical limitations also.

      1. Airplanes are metal tubes. Ever try to make a call in an elevator? A singlewide trailer? It's difficult or impossible.
      2. Even if you could get a signal in a plane, you're several tens of thousand feet up. You can see dozens of cell towers but go into and out of their range very quickly at 600mph. Cell tower networks aren't designed for this.
      • Re: (Score:2, Insightful)

        by Anonymous Coward

        Then why were people able to make cell phone calls on the 9/11 planes with no problem. The couple I've heard were long and clear.

        • by xaxa (988988)

          I read that the calls made from the 9/11 planes were made using the satellite phones (the ones where you swipe your card in the handset integrated into the seat and get charged $20). BICBW.

      • So you want to say that you are not _allowed_ to use phones on the ariplanes because you _can't_ use them?

      • Re: (Score:2, Informative)

        by DaveV1.0 (203135)

        While you are correct about #2, #1 is demonstrably false.

        As far as the GPP goes, the reason cell phones were banned on airplanes was concern about possible interference with avionics and instrumentation. The ban goes back to when cell phones were first popularized.

      • by Trogre (513942)

        1. I'm sorry that is just not true. What you're describing is a Faraday Cage, which needs to be built rather carefully, and then is usually tuned to cancel only certain frequency ranges. A plane just doesn't block cell phone signals. Unless the ones I've been on have had repeaters somewhere in them.

        2. Agreed, if you're near a metropolitan area. Between cities you might see three or four, but that's just a guess.

      • Re: (Score:3, Insightful)

        by InakaBoyJoe (687694)

        Sigh, mythbusting time...

        1) The "metal tube" myth: Get in an elevator, and compare the performance of a 2G (GSM, CDMA) phone and a 3G (UMTS) phone -- you might be surprised. In the 2100MHz band at least, most 3G phones work just fine.

        2) The "hundreds of MPH" myth: Nope. Phones are not banned on high speed trains in Europe or Asia, which also travel at hundreds of MPH. The story I heard was that it's not the speed of the handoffs that's the problem, it's the fact that a phone in an airplane at cruising a

    • by ethanms (319039)

      1. Text messages work when voice calls are dropped for the same reason Morse can get through when SSB voice can't.

      At this point it's all data... so why would TXT get thru and not voice? only explanation might be that more data doesn't get thru, or that re-tries make it happen...

      but I don't buy the tone-signaling vs. voice argument... it's just bits...

      • by SuiteSisterMary (123932) <slebrun@noSPAm.gmail.com> on Tuesday October 07, 2008 @02:14PM (#25289655) Journal

        Lets say a cell tower has 64 voice channels available. Lets say there are sixty-four people on that cell tower holding conversations. Lets say somebody calls your cell. Ooops, no available voice channel; they get your voice mail. You get a 'new voicemail' notification through the dedicated signalling channel.

      • by ethanms (319039)

        Oops, I should have been more clear in what I wrote...

        I have issue w/ the analogy between morse code vs. voice for sms vs. voice... the idea being that morse code is more easily transmittable with questionable signal so it's intelligible when voice would not be, therefore sms would be better able to get through vs. a voice call.

        I understand the idea that a voice channel may not be available to place/receive a call on because of congestion. In fact I'd take it one step further to say that I'd probably have

      • by Vegeta99 (219501)

        You are correct, sir. Under GSM, a text message takes up exactly one frame. Voice is 50 frames per second. All the phone has to do is catch one frame (that the cell tower is going to keep on repeating if it thinks that your phone is on.).

    • by BLKMGK (34057)

      Well you managed to get one right pulling ideas out of your ass!

    • by debrain (29228)

      3. You're not allowed to use phones on airplanes because of paranoid ignoramuses and the insightful people who realize how bad it could get when people in a flying bomb know what's going on (and how annoying cell phones are).

      Airplane Instrument Landing Systems [wikipedia.org] use radio frequencies that I understand mobile phones interfere with. This is particularly a problem when the plane is far from a cell-phone tower (as mobile phones will increase their signal strength to reach distant towers).

      This has become less of a

      • GPS with WAAS will replace the glidescope. It's computed on the fly.

        • by ptbarnett (159784)

          GPS with WAAS will replace the glidescope. It's computed on the fly.

          With enough satellites in view, GPS provides altitude. That can be (and has been) used to generate a glideslope indicator that is indistinguishable from one generated by an ILS (instrument landing system).

          However, GPS doesn't always provide sufficient accuracy for an approach to a 200-foot decision height (the minimum for a Category 1 approach). And it doesn't quickly (and clearly) indicate when the GPS signal is insufficient, either because of satellites in view, satellite geometry, atmospheric disto

          • WAAS will be sufficient for Class I ILS landings. LAAS (loacal area augmentation system), where they put a GPS correction station at the airport to provide local error correction over VHF. LAAS will permit Class III ILD landings.

    • by Dan East (318230)

      1. Text messages work when voice calls are dropped for the same reason Morse can get through when SSB voice can't.

      That's a bad analogy, and infers that the fundamental radio carrier / modulation is different for SMS than voice. Both are exactly the same thing fundamentally - data that are sent using the exact same mechanism. The difference is SMS is small, discreet and not real-time, whereas voice requires long-duration, continuous connectivity. Morse code has greater range because it is just a carrier wave that is not modulated with data. It is the modulation of the carrier wave that is so difficult to decode on a

    • by PitaBred (632671)

      1. That's not the only reason. SMS works on the signaling channels, which are never "used" except for a very short period of time. Voice calls work over the... wait for it... voice channels! Sometimes those channels fill up, but you can still get through on the signaling channels.

      2. RTFA... not purely based on the distance from the tower. It's also based on the number of phones in the vicinity that use the same bands and are fighting for signals, meaning they need to transmit a "I'm here!" type message

  • by instinct71 (1076915) on Tuesday October 07, 2008 @01:51PM (#25289321)
    How do mobile-phone servers distinguish between a switched off mobile phone and a one that is 'out of reach' of the mobile towers ? I never understood how I get those two different messages. What mechanism is used to differentiate between a switched off phone and a one that is out of reach ?
    • by Isvara (898928) on Tuesday October 07, 2008 @02:09PM (#25289589)

      Disclaimer: this is for GSM -- other network types may be similar, though.

      When a handset is turned on, it sends an IMSI* Attach message to the cellular network. When you turn it off, instead of immediately powering down it sends an IMSI Detach message to let the network know that it is no longer available.

      If you lose signal, or just take the battery out, the network doesn't know that the handset is unavailable. It sends out a paging message to the last cell it was known to be in, and eventually to the whole network before giving up and returning an 'unavailable' message.

      * Or TMSI if it has already been assigned a temporary ID to use instead of its IMSI.

    • When a handset switches off and you are within coverage, it will signal to the cellular network that it is turning off. Similarly, if you receive a call and press the End key to reject the call, it will send a "busy" signal to the network, which can be handled differently to the usual "not available/did not answer", depending on how your network profile is set up.
  • by sycodon (149926) on Tuesday October 07, 2008 @02:01PM (#25289475)

    1. Dial number, tower recieves signal and discards number.
    2. Dial again, tower connects and routes call around the world before connecting to the called number.
    3. Tower waits for conversation to begin and injects random noise, removes every third word, and then disconnects.

  • Have possibly managed to mention the iPhone more? Considering the market penetration, genericizing 'iPhone' to practically mean 'any old cell phone' is a tad premature...

    • "TidBITS is an award-winning electronic newsletter and web site dealing primarily with Apple Computer and Macintosh-related topics."

      Maybe you should find out a little bit about the web sites you criticize before you shoot off your mouth.

  • by SeNtM (965176) on Tuesday October 07, 2008 @02:29PM (#25289901) Homepage

    Stay tuned to Slashdot for our next featured article, "The Mysterious Wheel."

    When we will discover:
    1. What is a wheel?
    2. Why does a wheel roll?
    3. What magic has created such a device?
  • signal strength? (Score:2, Informative)

    by Anonymous Coward

    "Your GPS looks for special signals from satellites, and then compares the strength of those signals to triangulate your position."

    No, GPS doesn't use the signal strength to calculate your position, it uses the relative arrival time of time signals from the GPS signals.

    So can I trust the author to get anything else right?

  • by harrkev (623093) <kfmsdNO@SPAMharrelsonfamily.org> on Tuesday October 07, 2008 @03:01PM (#25290363) Homepage

    In TFA, the explanation of GPS is total BS. The person writing the article does not even have the faintest idea how real GPS works.

    Here is the real story:

    Unlike in the article, determining the GPS position does not use strength of the signal, but the timing of the signals along with a knowledge of exactly where the GPS satellites are.

    There are two types of data needed by a GPS: almanac and ephemeris. Almanac just gives the satellite's orbit. This stuff does not change, unless a satellite dies or the government changes the orbits for some reason. Given a rough location and time, the GPS can use the almanac data to know which satellites it should be looking for. This is why an older GPS may ask for the time, date, and state you are in when first turning it on. The GPS can figure out this stuff by itself, but it will take a few extra minutes.

    Ephemeris data, on the other hand, needs to be refreshed every hour or two, and pins the satellite's location down fine enough to be useful. This data is encoded on the GPS signal, and may take a couple of minutes to get (very slow data rate). This is why getting a lock can take some time when first turning on a GPS. If you turn off a GPS and then turn it on 30 minutes later (even if you traveled 100 miles in that time), then the GPS will get a fix in under a minute.

    The reason that phones can get a GPS lock almost instantly is that they can get the ephemeris data from the cell tower. It is true the cellular network can have a pretty good idea where the phone is even without the GPS, but that extra information does not help the phone's GPS at all.

    • by PitaBred (632671)

      That's funny... I know Google Maps on my Blackberry (T-Mobile) will find me by the tower I'm connected to first (and give an accuracy of about 1500 meters) before it starts tracking in on the GPS signal, if I get the GPS signal at all. Phones most definitely get more than ephemeris data from the cell towers, and the author's description had at least as many facts right as your post does.

      • by harrkev (623093)

        The cell network *CAN* get a rough idea of where you are without the GPS. I am just saying that it does not help the GPS much, other than providing current ephemeris data.

        For google maps, the cell phone location is apparently used as a fail-over when the GPS fails (or while waiting for it to start up).

  • A) Cell Phones are only licensed for ground mobile. Using them in the air is actually a crime.

    B) They can interfere with the navigational systems.

    C) It's not just cell phones.

    Here is some real world reports:
    http://www.airnig.co.uk/emi.htm [airnig.co.uk]

    Studies have been conducted on confiscated equipment. While there are a lot of variables, it can and has happened and has happened in repeatable tests.

  • ....TidBits doing an article on how web servers work and what happens when their URL appears on Slashdot?
  • by FrankDrebin (238464) on Tuesday October 07, 2008 @06:34PM (#25293081) Homepage

    I heard this second-hand, so take with a grain of salt.

    When a CDMA phone is idle, and the network supports it, the phone enters "slotted mode". Slotted mode is where the phone sleeps for a period of time (potentially quite long time -- several seconds), then wakes up to determine if anyone is calling it, then goes back to sleep until the next slot. Obviously, this feature is a key to very long battery life.

    Apparently a certain CDMA carrier with quite sparse network capacity in the rural areas, switches off slotted mode on long weekends. They found out that when everyone goes out of town, their network can't handle it. So they force all the cell phones to drain their batteries by switching off slotted mode. They found their customers are very upset when calls do not go through, but not upset if they have a dead battery.

    Sneaky if you ask me.

We don't know one millionth of one percent about anything.

Working...