Recent Posts: Martin Taylor

Just a couple of days after I posted my blog on "playing with a femto cell" comes the news that Vodafone has announced the launch of its femto cell product in the UK.  Under the terms of the trial NDA I can't comment on whether this is the same product that I have been using, but perhaps you can draw your own conclusions.

I mentioned before that I thought I'd be prepared to pay a purchase price of about £50 ($80) to get the benefits of better in-home coverage.  The price announced by Vodafone is a whopping £160 ($260), or you can apparently lease it for £5 ($8) per month - presumably with some contractual lock-in.  It seems they want to skim the market for those subscribers who have non-existing coverage at home and really want it badly.  But at that price point this is hardly going to be a mass market success.

A few weeks ago, I took a call from a market research firm asking if I would be interested in taking part in a trial of a new kind of "gateway" device that would improve cell phone coverage in my home.  Normally I hang up on calls from people I don't know as soon as I think they're trying to sell me something, but I wasn't going to miss this chance to try out a femto cell.  Mind you, they didn't make it easy for me -- the call center agent insisted I spell out every single word of my home address using the phonetic alphabet!

I live in a rural area where coverage is not very good.  My son claims the credit for getting us selected for the trial, because apparently he reported on our poor coverage to our wireless operator through some kind of Web survey.  We seem to be right on the cusp between two cells -- I have my phone set up to display the area code of the tower that the phone is registered with, and as I walk from my home office into the kitchen I can see this change as the phone roams from one tower to the other.  Signal strength is only one or two bars, we never get 3G service, and calls often get dropped if you move around the house.

I'm bound by quite a strict NDA so I can't reveal the vendor of the trial device, but I can say that it's one of the big well-known telecom equipment players.  Getting it up and running was simplicity itself -- plug in the power brick, connect to Ethernet and then wait (for up to an hour!) while it configures itself.  And once the LEDs had stopped flashing furiously, we had 5 bars of 3G on our phones.  Well, two of our phones, anyway.  You can apparently register up to four phones with the device, but since our children don't live at home any more, I'd just registered two numbers: mine and my wife's.  More on this in a moment.

Femto cells in the US have GPS receivers in them, and only work when they are located in an area where the service provider has spectrum.  This is quite inconvenient as you have to plug in a GPS antenna and position it where it can pick up a signal -- which means very close to a window.  There's no GPS in the device I'm testing, and I guess that's because my wireless service provider has nationwide coverage in the UK.  I'm planning to take the femto cell overseas on my next trip to see whether it will still work -- but I'm guessing that the system can tell from my IP address that I'm not in the UK and will disable my service.

So how well does it work?  When you're in range, coverage is great -- as you might expect.  I haven't been able to measure the exact download speed, but subjectively it feels as fast as a WiFi connection when you're browsing the Web or accessing email.  But the range is not very good.  We live in an old house with thick stone walls, and the frequencies used for 3G (1900/2100 MHz) clearly don't penetrate very well.  Two rooms away and you fall right back to the lousy coverage we had before.

When my son comes home for the weekend, he complains that he can't use his Blackberry at all anywhere near the femto cell.  It seems that the signal from it is so strong locally that his phone can't pick up the signal from the cell towers, and since his number is not registered with the femto service, his phone reverts to "emergency calls only" mode.  Friends with other kinds of phone have been able to make calls okay, but this does raise a question mark about the possible negative impact of femto cells on other users nearby.

As a condition of taking part in the trial, I've had to respond to a couple of surveys. The two key questions here are (a) Do you use your mobile phone more often at home now? and (b) How much would you be prepared to pay to purcase this device?  The answers I gave were (a) yes I do, and (b) about £50 ($80).

There are some interesting questions about market reaction to femto cells.  Some commentators have pointed out that by installing a femto cell, you are extending the wireless operator's radio access network at your own expense -- not just in paying for the equipment, but also in providing the broadband access that offloads your traffic from their backhaul network.  All true, of course.  But if you live in a poorly-served area, and you can get yourself great coverage in your home with a fairly modest outlay, then why not?

Having said this, though, I think that femto cells are a relatively short-lived phenomenon.  The increasing ubiquity of WiFi support on phones and WiFi in the home, and the inexorable migration of voice to the packet world, will eventually consign the femto cell to history.  In the meantime, they will probably enjoy a modest success.  And (sadly for wireline operators) they may well become another contributing factor to landline substitution by wireless. 

The recent announcement by Nokia and Skype about a partnership to ship a Skype client pre-loaded on the new Nokia N97 has caused more than a few ripples among the network operator community.  Reaction in the US has been muted (perhaps because Nokia's high-end smartphones don't have much market share there) but some European operators are said to be "fuming" and threatening not to offer the N97 unless Nokia removes the offending software.  In a similar vein, Deutsche Telekom has said that it may block the use of the Skype client on iPhone, citing a clause in its contract with subscribers that bans the use of VoIP and IM applications.

Skype is available as a download for a number of smartphones, including iPhone, but in general it only works over WiFi, not 3G.  For the time being, at least, would-be Skype users are not clogging up 3G airwaves in North America with bandwidth-hungry voice calls that would otherwise be using minutes from their normal calling plans.

In the current regulatory environment, there is nothing to stop wireless operators from discriminating against or even totally blocking VoIP calls made over their mobile broadband data services.  Judging by the initial reactions of wireless operators to mobile Skype, it seems likely that at least some of them will do just that if they perceive their traditional voice revenue base is coming under threat.

The one big exception to this is Verizon, who successfully bid for a major block of 700 MHz spectrum that came with "open access, open application" obligations.  When Verizon Wireless rolls out LTE service using this spectrum (in 2010?), over-the-top VoIP service providers should have the opportunity to offer mobile voice services without fear of discriminatory practices.  Subscribers with a data-only plan from Verizon Wireless may then be able to choose mobile VoIP services from a wide variety of over-the-top providers -- and this should offer an excellent opportunity for fixed network providers, who have customer relationships with their landline and fixed broadband subscribers, to play a serious role in the mobile voice services space.

Over-the-top mobile VoIP services are, of course, perfectly possible on existing 2.5G and 3G mobile data technologies such as GRPS, EDGE and EV-DO, but offering such services with a guarantee of non-discrimination may require cooperation with the mobile data service provider.  Zer01, a recent new entrant offering flat rate unlimited mobile voice and data for $70 a month, appears to be following this model.  They don't name the mobile operator whose network their service is running over, but comparision of the Zer01 coverage map with AT&T's coverage map (see the Data tab on this map) makes it completely obvious who their partner is.

The promise of mobile VoIP is yet another reason why fixed network operators need to modernize their legacy voice networks and get ready for VoIP.  Separation of services from pipes is starting to play a big role in POTS attrition, as landline subscribers increasingly take up over-the-top VoIP services from the likes of Vonage.  As mobile broadband data networks become more open, fixed network operators will at last have the opportunity to take the competitive fight into the mobile services space.

I was lucky enough to spend a couple of days last week in Barcelona at the Mobile World Congress and I thought I'd share a few thoughts that came to me during the (very interesting) conference and keynote sessions that I attended.

Everyone that spoke (and I mean everyone - even Microsoft's Steve Ballmer) paid tribute to the iPhone as a game changing device.  "It's all about the apps" seemed to be the theme here.  Ralph de la Vega of AT&T Mobility came out with the startling statistic that 50% of iPhone users in the US download a new application at least once a week.  Presumably this is because the novelty of most of these apps wears off in a couple of days. Like this one for example.

While the iPhone is held up as an excellent example of the new role of the mobile phone as an "open" mobile computing platform, there were some contrarian views expressed.

One view was that, while there's a definitely a market for a phone that you can download lots of apps to, there's a much bigger market for a phone that comes ready to use with all the really useful apps already integrated.  And "the really useful apps" are defined to be Facebook, Windows Live Messenger, Skype and social email, at least according to INQ Mobile, the developers of this device.  It won the 2009 Global Mobile Award for Best Handset, so this must make sense to at least some folks within the wireless industry.

Perhaps the INQ Mobile product won this award because it was so well in tune with one of the key themes I kept running into, namely that "social networking is the next big thing for mobile phones".  This was nicely illustrated by a three-hour seminar on the Rich Communications Suite, a set of SIP-based standards being introduced by the GSM Association that covers Enhanced Phonebook, Enhanced Messaging and Enriched Calls.  There's some quite interesting stuff here, for example being able to see the current presence status of the contacts in your phonebook, and being able to share files or video streams while on a voice call.  But the pitch used to illustrate the value of RCS focused exclusively (and relentlessly) on its use as a mobile social network, and all the actors in all the examples were 16-24 year olds.  There are actually some very useful business applications for this kind of service.  But business applications did not get a look in.  It's as though mobile service providers are betting on all their revenue growth coming exclusively from the youth market.

Here are a few other views that I heard more than once:

• Mobile operating systems and software platforms keep proliferating, and this can't be good for the market.
• There are 4 billion mobile phones on the planet, so the potential market for apps is HUMUNGOUS (never mind that only a tiny proportion of these are smart phones that you can install apps on).
• The iPhone is brilliant but it still only has 1% market share, so please stop BANGING ON ABOUT IT.
• And finally: the wireless industry is incredibly successful and will be an engine of recovery for the battered world economy. There's a ray of hope after all.

Over the last year or so, I've started to see some innovative new kinds of home phone devices come onto the market, such as the Embarq eGo phone and the AT&T Home Manager.  So I was very interested to see whether there would be more of the same on view at the recent Consumer Electronics Show in Vegas.  And sure enough, there were several new products addressing the same general space -- enough to suggest that we're looking at a whole new product category here.

First prize for good looks goes to the OpenFrame 2.0 from OpenPeak.  This sleek and stylish device has been dubbed "the iPhone of home phones" and looking at the pictures you can see why.  While I was actually writing this blog, Verizon Wireless just issued a press release headed "Verizon Wireless re-imagines the home phone" and announcing the "Verizon Hub" which has been built for them by OpenPeak.  Yes, you read that correctly: Verizon Wireless.

Not as pretty, but interesting for other reasons is the NIMble Concept Desk Phone from Touch Revolution.  The notable thing about this product is the use of Android, the open source operating system championed by Google for mobile phones.  Android promises a nice open development environment and potentially offers us the prospect of creating MetaSwitch-specific telephony apps for this kind of device, supporting things like visual voicemail and access to phone service settings like call forwarding and find me / follow me.

The iriver WAVE-HOME is another good-looking device in very much the same vein.  This one features a camera that supports video telephony, and offers a range of multimedia services.  Although not apparently as open as the NIMble product, the WAVE-HOME does have a "Widget" based development environment.

With price tags in the $200-400 range, these devices are comparable in cost with the new generation of low-end "netbook" PCs based on the Intel Atom processor, and are obviously a whole lot less versatile.  So who would want one, and why?

There's one school of thought that these devices might just encourage households that don't own a computer and don't have broadband to take a very basic broadband service just to support the tablet / phone.  The device would allow them to do a limited amount of stuff on the Internet like reading news, weather forecasts, TV schedules and so on without having to learn "computing".  Sold with a second voice line (with a VoIP service) and with a very basic broadband service priced right (i.e. very low) it's just possible that this might win over a few converts.

But I can't help thinking that the majority of those attracted by these devices are already broadband-connected and computer-literate, and just like the idea of owning one of these things because they are cool.  I for one would seriously consider keeping the family's phone contact list on one of these things.  And I can also see myself setting it up as a digital photo frame when not being used for anything else.

Mobile phones have long been "objects of desire" and "lifestyle accessories".  This new generation of fixed network devices brings the home phone into those categories.  Which has to be a good thing for wireline service providers.

The NGN / IMS Forum has just published its latest "report card" on the status of IMS, setting out to debunk some of the myths and misinformation that have grown up around this technology.  The "myths" include such items as "IMS is too complex" and "there is no business case for IMS".

The myths about IMS all have some basis in truth, of course.  IMS is complex, but given the scope of the problem that it sets out to address, this does not come as any surprise.  The important question is whether it turns out to be complex to build real IMS networks in practice.  And there's no fundamental reason why it should be.  Smart vendors ought to be able to take the complexity out of deploying and managing IMS, so the question we should be asking is "how smart are the IMS vendors?".  I think you can probably guess my opinion on that one.

It's certainly hard to build a business case to deploy IMS.  In my view, all attempts to justify bringing IMS into the network to support some specific new application are doomed to fail, because it will always be cheaper to create a new application silo than to re-architect the whole network to make application siloes a thing of the past.  IMS requires a high level strategic decision.  You deploy it because you just know that it's the right thing to do.  That takes a lot of courage, and perhaps not many CTOs will want to put their heads above this particular parapet - especially not now, when every investment proposal is going to come under extra close scrutiny.

I think that that IMS has been following the curve of hype and disillusionment that is typical of any significant new technology; having been heavily hyped in 2006/07, IMS has now reached the "trough of disillusionment".  The curve of course predicts that IMS will shortly start climbing the "slope of enlightenment" before reaching the "plateau of productivity" over the next 2-3 years.

We see IMS steadily gaining traction, particularly with the larger service providers that we talk to, so we have reason to be cautiously optimistic for its future prospects. That doesn't mean IMS will be right for every network.  It's a lot easier to build the case for IMS if you have both fixed and mobile operations.  If you think that IMS stands for "I Must Spend" and that you'll never see a hard return on your investment, you may well be right - in the context of your own network.  But IMS, or something like it, does have the potential to lower operational costs and simplify the roll-out of new services.  It remains to be seen how many are prepared to put their faith in that potential.  In my view, the report card from the NGN / IMS Forum doesn't do very much to justify it.

I don't follow a large number of technology blogs, but one I do read regularly is Telco 2.0.  A recent post there entitled "Voice telephony - death or glory?" was bound to catch my eye, and I really sat up and started paying attention when it talked about a startup that "exquisitely demonstrates that the value is in integration of telephony and the Web".

The startup in question is called Fonolo, and their product concept is pretty neat.  It's aimed at consumers who regularly call large companies (such as airlines) where they are inevitably faced with deep IVR menu structures that force them to listen to multiple menus and press the right keys in sequence to get them to the service they are interested in.  What Fonolo does is to catalog the "tree" of menu options, show this tree on their web site, and enable you to enter the tree at any point.  You click on the point where you want to go, then Fonolo dials the call center, plays out the correct sequence of DTMF tones to traverse the menu tree, and then calls your phone, so that when you pick up you are right where you want to be.

The Telco 2.0 blog holds this up as an example of what they call the "2-sided business model", a concept whereby telcos add value to the interactions between parties who use telephony to contact one another instead of just acting as a passive pipe. In the case of Fonolo, value is being provided to consumers in saving them time when they are transacting business over the phone with call centers, and at the same time is providing value to call centers in promoting their use and making them more friendly to users.

I'm not sure that the 2-sided business model concept is a fundamentally new one to telcos, but I do think it provides a useful source of inspiration and new ideas for telco business transformation.  For example the Fonolo concept illustrates the potential for telcos to help their call center customers by providing direct entry points into IVR menu trees.  Cataloging these trees and using DTMF tone sequences to dial through IVR menus seems like a very clunky way to solve the problem.  Why not encourage call centers to assign a direct dial-in number to each entry point, and then, rather than publishing a long list of these numbers, provide the call center with Web-based "Call Me" buttons that can be embedded on the call center's Web site?

We've long believed that integration of Web and telephony is an important direction for the industry, and it's good to see others engaging in insightful analysis around this topic - and coming to the same conclusion.  This gives us further encouragement as we continue to invest heavily in mashup APIs and other aspects of Web / telephony integration.

Green has become a hot topic in telecom, and both telcos and equipment vendors are quickly climbing onto the energy efficiency bandwagon.  Verizon started the ball rolling back in June, when it published energy efficiency specifications that it expects all new equipment to meet from Jan 09.  Shortly after that, AT&T announced it was joining the "Green Grid".  There's even a Green Telecom conference taking place shortly.

Heavy Reading has just published an article on the energy consumption implications of migrating TDM networks to IP.  This claims that migrating a 10,000 line TDM switch to an all-IP solution would reduce power consumption from 925,000 kWh per year to 102,000 kWh per year, saving more than $70k per year.

It would certainly be possible to hold this up as a fine example of how telcos can contribute to the reduction of global CO2 emissions, but as with so many "saving the planet" stories, it pays to look at the whole picture.  It turns out that much of the reduction in power savings comes from decommissioning the line frames, which drive POTS lines and perform analog-to-digital conversion.  But this function is still needed somewhere in an all-IP voice network - either in a broadband loop carrier, or a device on the customer premises such as an ATA.  These devices typically consume in the range of 3.5W to 5W per port, equivalent to 307,000 kWh - 438,000 kWh per year for a 10,000 line switch.

Based on this (rather simple) analysis, it seems that moving to a next-gen solution for telephony is more energy efficient overall - but the green benefit is nowhere near as great as it might first appear.  Nevertheless, there's a very useful overall cost saving, and this can make a big difference to the business case for migrating a legacy TDM switch to IP.

One interesting conclusion is that a consumer VoIP service, where the phone is connected to an ATA in the home, would save the service provider $3-4 per year (per line) in lower energy costs, when compared with POTS delivered from a broadband loop carrier.  This energy cost is, of course, just moved to the subscriber.  Not a great saving to the service provider perhaps, but every little helps!