Blog Post

Who Needs Femtocells If We Have Wi-Fi?

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!
Samsung's CDMA Femtocell

Femtocells, micro-base stations placed inside the home to improve cellular coverage, are supposed to be the answer to operators’ bandwidth constraints. They’re also a new source of revenue for carriers and the startups and large equipment-makers who are building the devices. But so far, the market has failed to materialize, not least because consumers don’t want to pay a monthly fee or buy equipment in order to help carriers improve their networks. But as Wi-Fi gets embedded on phones and hotspots proliferate, are femtocells even necessary?

The question struck me after reading about China Unicom — the Chinese carrier that’s now infamous for offering the iPhone (s aapl) sans Wi-Fi — announcing that it would offer a femtocell product for its subscribers. ABI Research analyst Aditya Kaul notes in a blog this morning that, since first-generation iPhones won’t have Wi-Fi, the China Unicom femtocell business may get a boost from customers who are eager to surf:

The interesting bit is that China Unicom recently announced their iPhone launch. Was this meant to be timed with their femtocell launch? The fact that the initial shipments of the iPhone will not carry WiFi or even the Chinese equivalent WAPI, bodes well for their ‘3G Inn’ service.

On the enterprise side, we’ve seen Wi-Fi take on femtocells and win, and my gut tells me that will happen in the home as well. Using your own wired backhaul to provide Internet access off the 3G network makes sense, rather than paying a carrier to use your wired backhaul to improve the 3G coverage that will count against your data plan.

Would carriers try to turn the clock back on hotspot coverage in hopes of pushing their femtocell offerings? That would mean fewer phones with Wi-Fi, no more carrier-supported hotspots and, for carriers, some serious network re-engineering. My guess is that such an approach may fly in China, but not in most other countries. Wi-Fi use for mobile data access is now widespread, and so far most femtocells in the U.S. are voice focused rather than data focused. It seems to me that the femtocell market is still a solution in search of a problem (GigaOM Pro subscription required). That’s bad news for Qualcomm, Samsung, Airvana, Ubiquisys, PicoChip and others betting on the market.

35 Responses to “Who Needs Femtocells If We Have Wi-Fi?”

  1. I have summarized the WiFi vs. Femtocell debate here a while back. Check it out:

    Hopefully, it makes it clear why WiFi is not a “Femto killer” and Femto is not a WiFi killer. It is not all black and white. Clearly, SPs would not be deploying Femtocells if they did not see value in them. And customers are indeed paying for Femtocells, so they see value too even if they have WiFi already!

    The roll-outs have been slow. We are talking about a major shift for wireless operators — all of a sudden the number of their cell towers is about to explode and they need a new level of self-install, radio tuning, etc. All new concepts for managing wireless infrastructure and not a trivial shift for massive IT organizations! Until self-activation is smooth, the SPs won’t push the APs full throttle. So, we won’t really know how big the femto will be until the vendors offer true plug-and-play solutions and SPs iron out all integration.

  2. mikeeeee

    you guys don’t get it at all.

    the benefit either with femtocells or UMA is so cost effective it’s crazy.

    just in the knowledge that you’re hooked up indoors.

    i live in an area where i have no bars and have to drive 2 miles to a boatramp to make a wireless call.

    i can make wireless calls in my home courtesy of t-mobile [email protected]

    looks like a no brainer to me.

    get my voicemail at the boatramp or get live calls in my living room.

    i don’t think anyone if a UMA chipset costs 20 bucks more a phone that works out to less than a buck a month over 2 years.

    if a femtocell router costs 250 bucks that works out to 10 bucks a month.

    it still hooks you up indoors.

    the only people who have a visible downside staring them in the face are the wired telcos.

    you know, the nice mega corporations that want to be you sole communications source and just give you what they want instead of all that’s available.

    instead of controversy, let’s figure out a way to make it ubiquitous so everybody can go to the dance.

    i will buy 2 of the first androids t-mobile markets or 2 verizon droids and a verizon network extender when my t-mobile contract runs out in about a year.

  3. Anonymous

    UMA & Femtocell are both aiming at solving 2 major issues,
    – Lack of coverage
    – Higher Throughput

    The approach is user friendly with a seamless integration in the phone that also provides handover between ‘homecell’ and macrocell networks. SIP clients & Skype like solutions have major drawbacks here as the user need to launch a particular client and do not provide handover facility.

    The problem with UMA is cost as not only an additional Wifi chipset is needed but also a specific UMA client instead of the regular one. The cost is significant as it requires a special hardware design (I would assume the WiFi chipset per se is not the major cost but rather how to integrate it into the phone and avoid radio interferences) and also software licenses associated with the UMA client.

    On the other hand Femtocell as no impact on the phone .

    I believe it is better and cheaper to add a WiFi chipset in the femtocell in order to provide both 3G signal & Wifi that going the other way around (The UMA way).

    PS: If anyone has info on UMA cost (Software client) and any estimate on how much it cost to add Wfi to phone, feel free to drop me an email. I am very interested.

  4. To me, this is more than just offloading networks or having a better solution for voice than WiFi, this is about owning the gateway in the home. As homes become wireless hubs for everything from energy monitoring/control to remote healthcare, there will be an enourmous opportunity to aggregate all of this wireless traffic inlcuding the normal voice/data traffic. M2M traffic could become more prevalent than person-to-person traffic in the home over the next 5 years. Some of this traffic will lend itself to 4G networks moreso than WiFi, especially for remote objects that aren’t on local sensor nets (like Zigbee).

    The other unque opportunity for Femtocells is the fill the gap in the “continuum” between the home-car-office. Today, this is not well addressed by WiFi or even celular networks for that matter. Given femtocells can have an extended range well outside the home, this could create gap coverage (like accessing your media server from your car as your car is parked in the driveway or routing your health data to your trainer as your jogging through the neighborhood). Wifi cannot address this.

    Having said this, I still think the carriers should subsidize the fentocells giving the fact it allows them to address the 25% of homes that do not have a fixed line phone today and also offload their networks, More importantly, it gives them a gateway into the home. I would love to see a 4G/WiFi gateway (which is available) offered by US carriers as a way to provide the best of both worlds.


  5. Debbie Greenstreet, Texas Instruments

    Certainly there are some viable voice solutions available today over WiFi largely due to the fact that 802.11 in the premises is so prevalent. However, without protocols like UMA that require capability both in the handset and the network, you won’t achieve the seamless handover that you would expect with a macro to femto relationship. Despite today’s WLAN market presence, I wouldn’t give up on femto cells. Due to all of the hype, it may seem like the femto market is not taking off, but it really in its early deployment stage. Unlike WLAN, it dove into the residential market before being fleshed out and widely accepted in the enterprise/campus sector – and I believe the enterprise femtocell and picocell market represent significant opportunities over the next several years. As far as the operators’ relationship with femtocells; well yes, they do need femtocells for multiple reasons and they want to control revenue, but they also know they have to find ways to entice and satisfy the subscribers. Not all operators charge for the box. This summer the Vodaphone device was announced and is provided gratis to those currently subscribing at a given monthly rate. We expect to see new applications and conveniences (imagine the ‘home femtozone’ where your phone knows “home” and can now interface to other devices and applications) being offered as the industry evolves. Early femto applications in Europe already include the ability for users to track their children entering and leaving school. I suspect that someday, the services offered by femtocell solutions will be so prevalent we may take them for granted; just like we have come to assume with WLAN presence.

    • Debbie agree with the sentiment of your post – it is too early to write off femto. However the signs are not looking good to me.

      The idea that femto is justified because of some “home” service seems a bit lame to me. Most “homes” already have WiFi and most other electronics (media server etc.) in the home are WiFi enabled.

      Basically femto was designed to complement coverage and imho will find a place an easy to install and maintain pico (i.e. to fill gaps in coverage). Do not see how it can replace WiFi in homes (esp. if the primary benefit is to provide fast data speeds).

      • But it is not that most homes have WiFi. Sure they do.
        But most *phones* don’t.

        Perhaps 10% of the phones sold this year have WiFi.

        Most of those don’t have software to make it easy to use; many of those (as the person above said) have WiFi turned off to save power.

        So a box that makes 100% of the phones work better (not a small fraction of a small fraction) is a good thing.

        Voice is still the killer app for cellphones, and there are a lot of people using Facebook etc on smartphones who don’t care if it is WiFi or 3G just so long as it works.

        Few technologies “replace” another (WiFi didn’t replace Ethernet; WiFi and Bluettoth both sell a lot of chips every year)

        Perhaps WiFi is the perfect solution to iPhone users (perhaps…!) — but for everyone else WiFi is not ideal – wheras a femto might be

  6. mikeeeee

    to all you skeptics poo pooing UMA…..

    go scratch.

    i have had both t-mobile varieties, [email protected] and plain old [email protected] for well over a year now (this post is going through my linksys/t-mo router) they rock!

    wait ’til magenta wakes up and offers an ANDROID UMA device.

    hang on, my g-1 does everything but, call right now over wifi.

    just gimme a UMA app for my g-1 or a new ANDROID with it baked in.

    ps; i’ve shopped the other companies for their femtocell offerings and they’re clueless and overpriced.

  7. i could see femtocells take off eventually but the audience that would go for them would be people who have eliminated there Cable/DSL in favor of an all mobile 3G/4G internet environment. the carrier would have to provide the back haul. this could be done with a rooftop device that connects back to the call carriers via a microwave link.

    but the idea that people would move carrier traffic they are paying for over to another cable or DSL link they are paying for is ridiculous.

  8. phil swenson

    If you are saying femtocells are silly because carriers can just route calls over WiFi, I agree. This is what AT&T *SHOULD* do instead of their ridiculously slow roll-out of the microcell. But unfortunately, they aren’t so I’ll have to buy the microcell. Looking forward to the day I can make a call from my house on my iPhone.

    AT&T why do you suck so badly?

  9. Rita Conaty

    I’d love a femto. When can I have one?

    Of the carriers, T-Mobile is the only one with UMA and it is hardly a good advert: I’m not surprised they don’t promote it: I wouldn’t boast about it either if I were them…. I’m guessing the others looked at it and saw how poor it was which is why they don’t offer (it is a competive market after all).

    Indeed, as the post below says, do any other carriers in the world offer it?

    You can only use it on a *tiny* range of handsets: great if you have a Blackberry which supports it, but I don’t, nor do my kids. The choice is limitted.

    It isn’t even all phones which have WiFi – and most of them don’t.

    Great if you want to leave WiFi on – but I don’t as it kills the battery. So that’s useless service.

    Quality was poor and it cut in and out.

    Where I live, coverage is poor. If I could get decent coverage to my existing phone, so people calling me actually reached me instead of voicemail – I would pay for that. If I got email to my phone in the evening, without going to the PC, so much the better.

    Sure, in an ideal world “the carrier should pay”. But that’s like saying “the government should pay” – the only way my carrier pays for anything is through fees – and I bet putting in a femto to serve my house is cheaper than them upgrading the whole sate and my hoping I get some of the benefit.

    • amerikaaskop


      The others don’t offer it, not because it’s a poor solution, but because it is compatible with existing hardware and is free. No money to be made there.
      Even T-mobile offers an unlimited bucket and ‘forgets’ to make clear that you can use it without that $9.99 plan as well, which convinces many (see poster above commenting on my earlier comment) that a $9.99 plan add-on is required.

      I prefer the WiFi solution, because it will work anywhere, not just at home where your femtocell is. For international travelers it really is an amazing deal to call and use data using regular minutes.

      BTW, I have never had any problems with WiFi calls; as long as the WiFi connection is of sufficient speed there is no discernible difference between GSM or uma calls.

  10. Stacey

    It is often fun to have “religious wars” of different technologies, and maybe WiFi vs femto is a battle about to happen, but I think it is a bit more interesting than that.

    Sure, there are a lot of cases where WiFi and femtocells are alternatives, and can deliver similar benefits to users, but I’d suggest that each has its own advantages. As such, the situation is a bit more interesting than you suggest, and the opportunities for femtocells are rather rosier.

    Femtocells are still new, so saying “the market has failed to materialize” seems an oddly harsh judgement. It is certainly clear that carriers are now starting to launch service: Vodafone this summer, AT&T starting their soft-launch last month and others such as Unicomm now. These are 3G: both Sprint & Verizon started with 1X products but have said they’ll soon launch 3G.

    You quote ABI: they and other analysts seem to agree there will be a solid market.

    Femtocells can deliver a host of benefits to both customers and carriers: better coverage, increased data capacity and new services. Some of these could also be delivered by WiFi, but others can’t.

    Most of all, femtocells deliver those benefits to absolutely any cellphone: Bob Metcalfe famously said “Always bet on the installed base” and 3+billion cellphones is a pretty good start….

    Of course, there are applications that are better suited to WiFi: I prefer steaming video to my laptop than on a phone screen, for example. Conversely, for voice, or quick email, I prefer to have my handset to putting a laptop in my pocket, and waiting for Windows to wake up…

    I think most people see the two as complementary: different ways of delivering different services to primarily different devices.

    But that said, you might pose other questions: Why have so few carriers have suceeded with voice-over-WiFi offerings to date; why have so many people have bought 3G dongles and data cards in the last few years; or, if WiFi on its own were good enough, why do we get excited about the iPhone rather than the iPod Touch ?

    The FemtoForum is hosting a conference in San Diego in two weeks: I would love to invite you. Perhaps we could arrange to meet & discuss things there?

    You might be interested in some of the demonstrations of services and update on the many carriers deploying or about to launch.

    Rupert Baines
    Chairman: WG1 – FemtoForum
    Vice-President Marketing, picoChip

    • RE: “Conversely, for voice, or quick email, I prefer to have my handset to putting a laptop in my pocket, and waiting for Windows to wake up…”

      There are plenty of people who use cell/UMA over Wifi for voice over Wifi….and plenty use (your example) of a iPod touch and Wifi or Novatel’s pocket device.

      I don’t think anyone questions the *technical* merits of Femtocells – the insulting part is the carriers charging for the device as well as additional monthly charges to provide backhaul for them through their own connection. There’s a specific use-case for people with poor coverage in their own homes (as you can see in these comments) — but how big is that market?

      • TimB

        Sure, and I’m not saying femto is the only solution.

        But I am saying that WiFi is only fitted on a portion of handsets so for the other portion a different technology is needed. Given that is 90+% of the market, it seems worthy of note….

        It is not just voice coverage: as people use more data, femtos are a good way to deliver that: 60% of cellular data is used indoors – which presumably indicates WiFi is not solving people’s problems just yet.

        I think the charging model is something we will see different approaches to, which is what you’d expect in competitive markets. Certainly several carriers have femto pricing quite similar to the UMA ones discussed above: free if you use your minutes, or a flat fee for unlimited minutes. There are carriers with more aggressive offers too.

  11. @TimB – I think carriers do not like to openly promote such devices because they are usually used for customers who do not have acceptable voice reception. And the carriers don’t want that message to be too prominent, IMO.

    Stacey – IMO, femtocells are more reliable (I don’t have one, but know folks who do) for voice than wifi – and that is the primary reason for the device in the US – to give customers with poor reception in their homes better voice coverage. As you stated, data is an after thought. I have not had good enough success with Skype on the iPhone for me to use it regularly, so I would certainly consider a femtocell for home office coverage where I make 99% of my calls.

  12. T-mobile never used the femtocells, but instead equips many of their phones with wifi to enable them to make calls using uma. This is completely transparent once the connection with the WiFi router is set up. No subscription necessary, and no equipment to buy, since it works through any existing wireless router.

    Added bonus: works through any wifi connection, anywhere in the world. On my last trip to Europe I made calls to the US at US rates, without the need for a VOIP app like Skype on my phone. Simply tell the phone to give priority to WiFi uma connections.

    • Actually, it *does* require a subscription. The $9.99/mo [email protected] package. (This is different than the ‘at home phone’ they advertise)

      Problem is, this service seems to have disappeared from their new rate plans. Seems that you can still use UMA but it comes out of your minutes. Others are reporting that if you don’t have the $9.99 unlimited @home plan that it’s not working at all.

      I have the unlimited hotspot calling feature at home, and you’re right, it’s awesome. But if I had to pay for a femtocell with unlimited airtime when using it, I wouldn’t mind that either. If I have to pay for the femtocell to compensate for crappy coverage, AND it uses my minutes up, that’s a problem. AT&T’s rumored pricing is $19.99/mo per *account* (not line level – account level) and the airtime/data is unlimited. I would be happy with that.

      • PaloAltoWorldView

        No, T-Mobile USA does NOT require any form of subscription for use of UMA. There is a $9.99/month plan available IF you’re not otherwise on an unlimited plan or don’t want the UMA-minutes to be subtracted from your “regular” cell plan. So if you have an unlimited cell calling plan, for example — it wouldn’t make any sense for you to add the $9.99/month thing.

  13. I’ve always wondered why T-mobile’s Wifi/UMA combination wasn’t much more popular. It’s not prominently on their website, and they only have a couple UMA capable phones.

    • Stacey Higginbotham

      I wonder if it was perhaps ahead of its time. Only now is the mass market paying attention to issues such as getting good data quality on handsets and avoiding roaming charges using VoIP, where UMA shines.

    • Stacey Higginbotham

      I wonder if it was perhaps ahead of its time. Only now is the mass market paying attention to issues such as getting good data quality on handsets and avoiding roaming charges using VoIP, where UMA shines.

    • I use a BlackBerry 8900 with UMA. It’s such an excellent feature. I love that it helps me stay accessible while I’m out of the country.

      A good friend of mine uses it because it’s his best option. His house isn’t covered well by any of the “big four”. UMA takes care of that issue for him.

      My dream phone is an Android 2.0 model with a great keyboard, a strong processor, and UMA. C’mon T-Mobile! Make it happen!!!