Blog Post

iPhone Tethering? Be Careful What You Wish For!

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

It’s rumored that AT&T (s T) may soon allow people to tether their iPhones to their portable computers and use them as 3G modems for an extra $10 a month. Now that’s a wonderful idea — in theory. If it does indeed happen, though, you can expect one thing — your 3G iPhone may behave like a Ferrari with flat tires. Like an overloaded mule walking up a mountain trail, AT&T’s 3G network is already buckling under the strain of its many iPhone users.

You know what I’m talking about: at least once a day I get blue in the face because the network is so slow. And now imagine if hundreds of thousands of iPhone users signed up to use iPhone as a way to connect to the Internet and surf the web, check emails, watch videos and swap big files. You would have a lot of traffic flowing through AT&T’s pipes. “So what?” you might think, “AT&T has a 3G network that’s supposed to be fast.” Not quite!

Just because we have 3G phones (and networks) doesn’t mean we can get higher speeds. That’s just not the case. Most of us connect to a cell-tower, which in turn has a base station that’s connected to the phone company’s network. This connection between a base station and the wireless operator’s network is over (multiple) T-1 lines. Bandwidth coming into a base station — anywhere between 1.5 Mbps to 7.5 Mbps — is shared by people who are connecting through that base station. And that’s not enough to support blazing fast speeds for all the current iPhone users, much less the new tethering users.

Sure, there are some who are experimenting with microwave technology and deploying a Gigabit Ethernet connections, but the fact of the matter is that the backhaul networks of today are woefully ill-equipped for a real wireless broadband future. That is one of the reasons why demand for wireless backhaul equipment is on an upswing. Research firm Infonetics predicts that phone companies around the world will spend $10 billion by 2011 to make sure you can get your 3G fix on an iPhone.

But all that is going to take sometime. Sure, tethering is a great idea, and I want it now. But I also want to get 3G speeds, and this effort may disappoint.

Related stories:

1. New iPhone will jumpstart wireless broadband
2. Is iPhone’s 3G connection a disappointment
3. 4G Wireless & the ensuing bandwidth boom

27 Responses to “iPhone Tethering? Be Careful What You Wish For!”

  1. Wamphry

    Call me part of the problem, but I jail-broke my iPhone a week after I bought it. I have been tethering on AT&T’s network for almost a year with decent speeds usually 1-2 mbps using a Cydia app called PDAnet. If my phone gets stolen, I have an app running in the background that tracks my phone as long as it is turned on. I also have a free voice nav gps driving app called xGPS that is FREE… completely FREE, and it’s nearly as good as my Garmin Nuvi! In fact I take that back, it’s better than my Garmin because it pulls map info from google maps and doesn’t cost $89 a year to update! AT&T want’s to charge users for tethering but, how can they when some of us have jail-broken phones and get it for free already. Until Apple and AT&T can squash the Jail-Breaking… you won’t see a tethering option for AT&T unless the just decide to give it as a free service!

  2. It is T’s burden for having proper backhaul for wireless customers — esp when in regions where they are the LEC.

    No point to tell people they can drive @ 80 mph when you only provide a two-lane highway since everyone end up getting stuck.

    There are microwave and E-Band microwave technologies and fiber lines they can deploy to improve backhaul consider how much financial resource they have.

    It is a bloody shame.

  3. daTruth

    Tethering WILL NOT place a burden on AT&T’s network. Your laptop while tethered, CANNOT access any more bandwidth than what your iPhone can.

    Any concerns from AT&T contrary to this are pure FUD just to make more $ off iPhone users.

    If AT&T does charge for tethering, it will be a bogus charge and customers should strongly voice their complaints against this.

  4. I really don’t think it will cause taht much of a strain to be honest. Rarely would i ever use the tethering feature, but when i need it, it needs to be there. Example of i’m up at the cabin and i need up update my site, a simple 1mb worth of data (comic of the day), or send an email with a file that’s on my computer, again probably won’t bog down the network. I highly doubt anyone will be playing online games like WoW on that connection, or downloading movies. It’s an important feature for a lot of business people, and they should allow it.

  5. It is hard to decided whether or not this is a good thing. On one hand, I would really like to add the feature to my service for ten dollars. On the other hand, I get such unreliable service in the Chicago area already. I can’t imagine my phone service quality will go up with so much bandwith being used. Maybe it will for ATT to upgrade their networks or offer tiered internet service or something. I would pay a little more for better ipone data service if there was a way to differentiate different customers. I mean we all pay more for 3G right?

    Matt Kennedy

  6. Scarhawk

    AT&T does not build out ahead of demand, they consider it a poor strategy for deploying capital. Unless customers are eager to pay and/or are screaming to get what they think they’re already paying for, money generates a better return in the bank than in the ground. If they had any real competition it might be a different story, but they don’t, so business-wise they’re making the right decision.

    Meanwhile, how hard is it to drop a metro fiber connection to a base station? If tall buildings can get backhaul, so can towers, somebody just has to string it up the hill.

  7. I am tethering now for almost a half a year on windows mobile. The WM5 worked at 2.3mpbs as a dial up connection via either a USB or bluetooth connection.

    Since the Ev-do upgrade for the T-mobile service and to WM6 on my smartphone (I used the MDA during dial up tests and Wing currently), the link is now displaying 10mbps when I open the connection icon in my task bar, though only connects close to 2.3 up and .5 down according to tests.

    The connection speed varies, it is good for loading pages and non-complex java programs, but streaming video still needs additional buffer time.

    A contemporary of mine is on the Sprint system using a Mogul and is connecting at slightly a faster rate.

    Regardless, tethering is a great alternative to searching for wi-fi signals. It takes only a few minutes to set up, actually, it is pretty much plug and play as they have updated the configuration to automatically install the proper drivers.

    Sprint offers web access for $15 a month according to my friend. I am paying $30 a month for T-mobile and have for the past three years now with mobile web capabilities.

    • Sorry, Dan, you’re not getting 2.3Mbs, though that might be what Windows Mobile tells you when you connect. That’s the speed of the local connection to your PC, not your actual bandwidth.

      The best wireless 3G download speeds are in the neighborhood of 700k-1Mb currently.

  8. Dropping By

    First of all, AT&T has to have a minimum of 4 T1s to a UMTS site to boast a 3.2Mbs advertising metric. Two, a draw back to UMTS R5 – R6 is an ATM backhaul for cell site facilities. This kills any efficiencies you could possibly gain on a T1 based or ethernet network in the transition from RNC to SGSN/MSC. R7 or R8 is to oliviate this draw back but this will kill AT&Ts TDM revenue stream from legacy POTS line or T1s.

    There is the reason data services are slow. A T1 on average is eight times as much as ethernet facilities. AT&T mobility leases millions of dollars of T1s from AT&T and with the lose of that revenue stream kills AT&Ts overall bottom line. If you don’t believe that AT&T is still charging AT&T Mobility(Cingular) for lease lines, check their quarterly reports for Q1 of 2006 and Q1 of 2008. There should be a huge shift in OPEX on AT&T mobilities part in the negative (good) and a huge drop in income on AT&T part.

    If AT&T mobility converted to ethernet today, which they can do with inexpensive cisco CPE equipment, AT&T would look live Verizon. Legacy costs and revenue for AT&T is a win-win because the equipment that the legacy facilities are on has depreciated to $0 and all revenue that comes in is straight cash for them.

    Telcos are full of it and it is not about public service but money, but how can they survive if they didn’t think this way. Maybe they would have needed a bailout. I like their arguement about not being obligated to rural carriers if the USF funds are not sufficient (“if it don’t make dollars, it don’t make sense”). The Telcos will not build out these areas if Obama doesn’t supply 100% of the capital needed or some rural competitor comes out smelling like roses and give them a reason to want central SD.

    For the iPhone argument, they are wanting to know how bad you want it. If the phone worked the first time, it would be too efficient and they need you to waste so money can be made off of it, “by design”.

    I’m getting an iPhone on the 30th of this month.

  9. AT&T literally makes me sick. The marketing guys there are saying, “it’s all about capacity utilization and ROA. Milk the damn system and fc*k customer satisfaction.” As long as there are a few big carriers, we will be stuck with companies thinking like the government. When will we ever get an all IP, open network? Maybe after India and China have one and keep kicking our asses.

  10. Yep…T’s problem is that they are *cheap*. Fact is that they have plentiful of b/w from their local business. Clearly they have backhaul to deploy U-Verse and DSL, they sure could manage to allocate capacity for their wireless play if they wanted to.

  11. When AT&T Wireless was Cingular, around 10 years ago, their network couldn’t support voice traffic for the accounts they sold, resulting in dropped calls and “person not available – try again later” messages for those trying to call Cingular phones. They didn’t listen to customer complaints, so they and SPRINT were sued by the FCC to stop them from selling more accounts until their networks could support them. When they became AT&T they increased prices (for call forwarding) in the middle of customer contracts and then refused to refund equipment costs for those who wouldn’t accept the changes. Then they wouldn’t return positive balances after customers disconnected until the customers threatened to sue.

    AT&T’s problems aren’t those of an early adopter (plenty of real early adopters in Japan, Korea and Europe). It is a company to be avoided where possible.

  12. Jacob Varghese


    While the iPhone makes up a smaller percentage of smartphones, the users make up a much larger percentage of mobile web usage.

    “Combined, the iPhone and iPod Touch represent 15.5 percent of all worldwide requests. In the U.S., the iPhone OS (if you include the iPod Touch) now accounts for 48 percent of the smartphone market.”

    I have an e71 and I am quite envious of the mobile web experience on the iPhone.

  13. Why is it I keep hearing claims that this tiny portion of users (yes, check the stats) can affect the entire network? If this tethering were allowed for Windows Mobile users which account for a MUCH larger group of users (yes, check the stats again) then I could understand concern about bringing a whole network down. Just because the iphone has more hype and more people yelling about it, does NOT mean there are more users.

  14. Why not take this as an opportunity to visit Romania, Bulgaria, Estonia, or a number of other European countries where 3G broadband has been working just fine for quite some time now.

  15. I am not missing the point, Om, notice the terminology used. :)

    Yes, it’s old-school at the moment, but improving it is not complicated. Motorola has a solution already –

    Besides, markets with 3G fully deployed can’t possibly be on T1s at this stage anymore.

    I remember the oversubscribed dialup lines from 1990s, so this is really nothing new. At first we had different tiers of providers. Competition forced everyone to upgrade to 56K V.90 evenfually. K56Flex was interesting in the meantime. But what most people may not know is how that was accomplished – ISPs upgraded to DS-1 termination equipment.

    Now we are witnessing the same change. Except this time it’s the DS-1 circuits that are no longer adequate for data (GSM is a rather thrifty codec in terms of voice quality, which is why you only need 13kbps to transmit it) and metro Ethernet is replacing it.

    In terms of profit, providing pure voice is a lot more profitable than having to provide all that high speed data. Unfortunately for them, we want more for less. :)

  16. Mike Cerm

    I don’t think that this is an AT&T problem. Sure, they have the most popular feature phone on the market, and it’s straining their network, but I don’t think their connection speeds are that much worse than Verizon and Sprint.

    The real problem is that we have 3 different, incompatible nation-wide networks, all of which are a decade behind the rest of the world. We have 3 companies slowly building out infrastructure, resulting in mediocre coverage and lackluster speeds no matter which network you’re on.

    It would be nice if the competition made things better, but it doesn’t. We basically have 3 monopolies, each with abusive pricing (text messaging?), and terrible service.

  17. Leonid

    you are missing the point — the back end of AT&T is still old school despite all the fancy 3G networks. Put more load on this and you will be surfing at WAP speed unless they improve their backhaul networks drastically

  18. I disagree.

    The option to have moderately fast Internet anywhere beats the slightly faster than a modem (EDGE) option. We are headed towards ubiquitous 3G, so AT&T really doesn’t have a choice but to improve backhaul network capacity.

    It may be painful at first, but such is the plight of early adopters. :)