Speaking Engagements & Private Workshops - Get Dean Bubley to present or chair your event

Need an experienced, provocative & influential telecoms keynote speaker, moderator/chair or workshop facilitator?
To see recent presentations, and discuss Dean Bubley's appearance at a specific event, click here

Friday, December 29, 2006

Predictions for 2007.....

1) Increased focus on manufacturers selling multiple "diverged" devices to users.

Imagine you are the CFO of Nokia, Apple, Motorola or similar company. What would you rather sell to your customers? A single, very complex product which risks a long time-to-market and significant integration problems - or two or three lower-priced, simpler devices which, if they're well-designed, should still have decent margins?

2) A lot of noise about VoIP over 3G.

Both "official" carrier-centric versions like VoIP over EV-DO Rev A, and "unofficial" versions using handset software clients which exploit flatrate data channels, SIP capability and clever codecs will be in the spotlight. An interesting dynamic may evolve here, as it looks like the CDMA world is far in advance in this area from carrier viewpoints, while even the cleverest startup's handset client in the UMTS world won't get around limitations of 3G coverage or limited uplink capacity before rollout of HSUPA.

3) Emergence of corporate-focused MVNOs

OK, I've been expecting this for 3 years, but maybe 2007 is when it really happens.... I get the sense that "proper" enterprise mobility needs a mix of skillsets possessed by few mobile operators unless they are integrated as part of a major national fixed/mobile player. The ability to mix basic mobile voice products, with mobile data devices, back-end server and application integration, indoor coverage solutions, and, absolutely critically, PBX integration (ie NOT just dogmatic and totally unrealistic "replacement" with mobile IP Centrex).

4) Continued uptake of various dual-mode services & handsets, but they won't change the world

I posted last week about UMA/non-UMA developments & my predictions. Overall, it's still very much a niche game - I'm willing to bet there won't be any "iconic" global-massmarket consumer handsets emerging that include WiFi. Outside the US, the true sign of maturity of a handset technology is its appearance in prepay-centric phones and service plans.

5) Spectrum lobbying noise, regulation momentum and lawsuits ratchet up several notches.

2.5GHz licences, "spectrum neutrality", refarming 900MHz GSM for UMTS, more low-power GSM licences, analog TV switch-offs, mobile TV spectrum auctions, ITU World Radio Conference.... it's going to be a busy year for the lawyers.

6) IMS confounds both its critics and its evangelists, but needs to improve integration ASAP.

No, it won't disappear completely or implode under its own weight. No, it won't be used for more than a tiny fraction of operator services worldwide by the end of the year. Yes there are still huge gaps in both standards and solution availability. But a few major mobile operators - particularly those in dirigiste markets that develop/specify their own handset software in detail - will launch proprietary almost-IMS. Others will use "small" - almost trial - IMS adjuncts to offer VoIP and other services to non-core devices like PCs. Fixed use of NGNs will increase substantially, although they generally won't be "full IMS", either.

The key lesson for IMS advocates to learn during 2007 will be integration - come down from your ivory towers & learn how to blend IMS with non-IMS - the real Internet, enterprise networks, SDPs, music & TV platforms and so forth. If the IMS community doesn't wholeheartedly embrace these areas of integration, in both the network and on devices, it will stagnate in 2008 and die in 2009. Isolation and "purity" is doom.

7) Navigation becomes rather more important on mobiles. Mobile search doesn't.

Although it will take some time, I do see handset-based navigation capabilities becoming more prevalent. Phones will start to ship with built-in GPS in small numbers, and unlike WiFi dual-mode or 3G, this will be seen as "cool". Generally, cool stuff in phones is not the service-oriented stuff - think cameras for local storage, MP3 players with memory cards, snazzy externals. I have my doubts about operator-centric location based services (except E-911 in the US, and a handful in Asia), as the exponential curve on memory and processing much faster than over-the-air transmission. You could get a detailed map of any country in Europe on less than 0.5GB of flash memory, just downloading updates or alerts when necessary. And it might have been my most controversial post this year, but I still expect that most "Mobile Search" companies will be selling their new 2007 collections principally for the Emperor's wardrobe.

8) The City WiFi bubble bursts

I've been amazed this year by the number of local authorities willing to spend yours & my hard-earned taxes on subsidising WiFi on lamp-posts, or at least permitting other operators rights-of-way to do it themselves. In my view, it's a completely over-hyped proposition that doesn't merit its "me too!" bandwagon status or breathless "Digital City" marketing. It's only useable for a handful of boring outdoor-only local services like CCTV cameras and connecting traffic wardens' handhelds. The killer: it won't work reliably indoors. It's like 3G but even worse. Next up on the list for over-funded local authorities to burn our money.... Municipal Fibre

9) Flat-rate data becomes the norm, with browsing the killer app, driven by high-res screens

I'm still waiting for my trial X-Series phone from 3, but I've been increasingly impressed with browsing experience recently. While cheap data tariffs are one critical driver, another has been largely overlooked - increasing screen resolution. The standard for mid-to-high end phones is now QVGA (320x240 pixels). This will increase, either with Nokia's weird 416x352 (or something like that) or more standardised full VGA (640x480). I'm a firm believer that there is no "Mobile Web", and that most people would much prefer a mobile broadband ISP experience, accessing the one, real, Internet. And, of course, that means their favourite web brands & downloadable add-on client software too. The signs are already there at the end of 2006, but 2007 will be the year the mobile industry stops fantasising about beating Google and Yahoo and Skype, and instead just gets on with optimising their performance for their customers. Long live the Smart Pipe strategy......

10) No, No, No, No, No

OK, this post is already long enough, so I don't have time to detail my reasons for all of these, but I'm sure they'll crop up on the blog in coming months. Mobile IM won't replace SMS (sorry VoIP fans...). Laptops with built-in HSDPA won't sell much (and even where they do, the cellular bit won't be activated by most owners). WiMAX will get a few more major operator advocates, but still won't be seen as a threat to "normal cellular". Mobile TV won't make much headway. Web 2.0 stuff like social networking really won't be a big deal in mobile outside Japan, Korea & maybe the US, unless carriers work out a way to give decent Internet access & capable devices to prepay users.

Oh, and maybe Apple's Phone-i (hey, Linksys got the iPhone brand....) won't play music at all, but will be "just a phone". See point 1.

Wednesday, December 27, 2006

Skype + Sony Mylo + T-Mo US: a big deal?

Andy Abramson over at VoIP Watch thinks that Skype is being seriously subversive with its tie-up with T-Mobile in the US, allowing hotspot-based voice calling at a flat rate.

I'm not that convinced at all.

Two main reasons:

1) Inbound calls
2) SMS

Plus myriad other minor reasons. Apart from anything else, maybe Starbucks in the US is different, but my local one here in London plays such awful piped background music that the last thing I'd do there is make hours of calls on my mobile to my friends or important clients.

Thursday, December 21, 2006

Dual-mode FMC services, 2006 Final Score: UMA 6, SIP 6

I've just done a quick count of launched operator services based around dual-mode WiFi/cellular handsets. Note that I 'm talking here about operator-based services, rather than the numerous options for DIY or startup wVoIP or ISP offerings.

I reckon the number of commercially-available UMA services is 6: BT (which has just launched its WiFi version as well as Bluetooth), T-Mobile US, Orange, Telecom Italia, Telia Denmark and Saunalahti. Orange has launched in 3 (I think) countries, though.

And the number of commercially available SIP-based dual-mode services is also 6: NTT DoCoMo & KDDI (both for enterprises, and sold with PBX partners), Deutsche Telekom, Arcor, NeufCegetel and Free/Iliad.

(note: I don't know much about the KDDI service, so conceivably it's proprietary rather than SIP, in which case my apologies)

In addition, there's numerous smaller SIP-based options from the Truphones & the Frings & Woizes (& their numerous brethren), and probably a few ISPs and VoIP providers that ship a smartphone client. Also, a bunch of individuals have independently downloaded Skype & assorted other softphones to their WiFi-enabled cellular PDAs and smartphones. There's also the Korean & Brazilian Bluetooth CTP services.

My prediction for 2007?

I reckon that by the end of the year, in terms of launched services, we'll be on:

11 UMA operators (some in more than one country, maybe including a couple of MVNOs as well). I'm excluding UMA used for non-dual-mode applications like femtocells.

30 major operators with SIP-based dual-mode services (probably about half of which will be enterprise-only).

There's likely to be some overlap, with operators like BT and France Telecom probably having both variants for different customer groups.

I'm currently working on updating my market forecasts for the handset numbers, but I'll be deeply surprised if more than 3m UMA phones ship in 2007. Blowing my own trumpet a bit here, but my original forecasts made 18 months ago, in June 2005, suggested 0.6m UMA phones would ship in 2006, and 2.4m in 2007. The '06 number - considered by many as horribly pessimistic at the time - now looks a bit too high, as BT and T-Mobile pushed their launches later in the year than I'd expected, but I still reckon the '07 figure looks pretty reasonable, probably with Orange & T-Mobile US (if it does a full rollout) accounting for most of the volume.

SIP-based dual-mode devices is a bit trickier, as there's a lot emerging from Nokia, HTC etc where the WiFi is primarily for data, but which can support voice given the right software. I suspect the number shipping to customers with VoIP preloaded is probably going to be 4m+ but I haven't done the numbers yet.

For both of these (and especially the SIP devices) there will also be a lot of dual-mode phones which ship - but which aren't actually used by dual-mode VoIP purposes by their owners. Whether they "count" or not will depend on whether you're selling phones, software licences or telephony services.... aah, the joy of market statistics.....

Monday, December 18, 2006

T-Mobile US UMA service... not exactly ready for national roll-out

Interesting article here about the trials & tribulations of the new T-Mo US dual-mode service. No wonder there's been no nationwide rollout in the US yet.

Even leaving aside the difficulties in getting the technology tuned... is it just me, or are other people also staggered by the line "regularly exceeded their 2,000-minute-per-month T-Mobile family plan". I know you lot the other side of the Atlantic talk an awful lot on the old dog-and-bone, but even so, that's an awful lot of chatter. I probably talk about 200 mins / month on my mobile (incoming & outgoing) plus maybe another 400 on Skype & fixed-line put together.... and that's including doing hour-long briefing conference calls.


OK, maybe 30 mins each per day isn't that much, if you spend all day at your desk or in your car making sales calls. And I guess it puts people in the rare category of actually regularly using the WiFi / cellular handover bit, if they're always gabbing away to clients on the phone when they get home in the car. I'll add that to my list of "niche target markets for dual-mode usage", along with Finns' second homes in remote forest areas which have better broadband connectivity than cellular coverage.

Thursday, December 14, 2006

UK 2GHz band.... so just what does "designated" mean?

As you've probably read elsewhere, Ofcom is auctioning off large chunks of prime spectrum in the 2.0-2.6GHz range.

Various commentators have already highlighted the likelihood of BT bidding for some spectrum for WiMAX use, as well as the likely wrangles over whether it should be preserved for 3G UMTS instead.

Now, theoretically, according to EU law, chunks of this spectrum is "designated" for 3G. But on the other hand, Ofcom has already been a leader in pushing "spectrum neutrality" and appears to want to make this a test case for much of the rest of Europe. Other European regulators are split on the issue - some are very much cellular-driven, others like the idea of neutrality on the grounds of competition. It seems probable that the immediate short-term beneficiaries will be lawyers.

I suspect that semantics may well come into play. There seems to me to be a clear distinction between the terms "designated for..." ; "designated exclusively for...." and "designated irrevocably & exclusively for...."

The ITU's definition of the verb "to designate" is less than helpful on the matter:
"To identify a set of characters that are to be represented, in some cases immediately and in others on the occurrence of a further control function, in a prescribed manner."

SMS over IP standard

Had an interesting call earlier with LogicaCMG, a company with a long heritage in SMS and MMS, and which has a fairly messaging-centric view of the world when talking about IP and IMS. Aside from its current pitch around converging different messaging types onto one platform, another topic which cropped up was a new standard agreed by 3GPP a week or so ago.

Called TS 24.341 this is basically an agreed standard for supporting receipt & transmission of SMS over an IP network - eg from a dual-mode phone connected in WiFi mode.

Now, being 3GPP, the whole thing is clearly very IMS-centric, involving the usual registration of a device with an alphabet soup of HSS and CSCFs. I haven't had a chance to wade through the whole thing, but it certainly looks like it's been designed with some specific problems in mind. What's not immediately obvious to me is whether the same approach (or a similar one) will work outside of IMS, so that (for example) a VoIP provider or an enterprise with an IP-PBX could also deploy an IP-SM Gateway.

As I mentioned a month ago, the whole VoIP+SMS issue is a critical one now, especially for the new breed of mobile VoIP providers. I hope they can use some aspects of the new standard to solve one of their more intractable problems - and I encourage both infrastructure suppliers like LogicaCMG and their handset SMS-client software counterparts to consider both operator- and non-operator use cases for SMS-over-IP in their product development roadmaps.

Tuesday, December 12, 2006

Bluetooth CTP... the original FMC solution, still lurking around

At the WiFi / Cellular conference I chaired in Dallas a week or so ago, I was rather surprised that one of the Korean presenters was interrogated about Bluetooth-based FMC. Not Bluetooth UMA as per BT Fusion, but the even earlier dual-mode variant called "cordless telephony profile" CTP, which was originally launched by Korea Telecom, eventually signing up 100k or so subscribers before disappearing quietly.

CTP was one of those good ideas that appeared a bit too early, suffering from a lack of interest in deploying Bluetooth access points, and an equal lack of interest by operators in phones which had a potentially-threatening Bluetooth profile which could cannibalise minutes.

So I was a bit surprised to find out (belatedly) that CTP has actually been granted a new lease of life in Brazil. Now, I can't really see it making a comeback on a much broader stage, but it just goes to show that there's always more ways to create a service than you might think.....

Normal service is resumed....

Apologies for the patchy posts over the past 2-3 weeks - a combination of moving to a new house, travel for a week or so, and getting a new broadband line sorted has soaked up a lot of time.

There's obviously a lot to catch up on, including 3's flatrate pricing, a bunch of chat around picocells, and an attempt at enterprise-grade UMA. I'll try & backfill as much of this as possible over the next few days.

Actually, on the broadband front it's interesting that my new gateway/WiFi router (a BT Home Hub) has no trouble blasting 802.11 through 3 ceilings from basement to bedroom at 5 bars' signal strength, giving my laptop 80% of the throughput I get on my wired-ethernet desktop. Anyone want to lend me a trial 3G femtocell to see if the inhome RF transmission is just as good?

One last "thought for the day".... GSM and its UMTS 3G siblings often like to bill themselves as "the world's most popular communication technology". Actually, although the majority of it is wired, I suspect that actually that accolade goes to Ethernet, which has apparently sold more than 8 billion ports during its life. It's also got breadth as well as volume, spanning from plain-old PC connections to WiFi and carrier backbones, and scaling up from 3Mbit/s in its first incarnation to 10Gbit/s over fibre today... it's probably what makes me think that WiMAX has more of a chance than some observers believe, especially given Intel's longstanding love of ethernet (and Microsoft's, and Cisco's....)

Tuesday, December 05, 2006

Attn Disruptive Analysis contacts: Change of phone #

Trying to coordinate my travel schedule, with the technicalities of simultaneously moving premises has unfortunately left a gap - despite my request for a forwarding announcement to be put on my old fixed telephone number (+44 20 7723 1259) to a new number, I understand this has not been actioned, and callers are getting a "number unobtainable" tone. I will rectify this as soon as I get back to the UK tomorrow, but in the meantime please use email or use my mobile number +44 7941 100016 to contact me if urgent.

Friday, December 01, 2006

Ericsson and Intel - IMS phones or IMS laptops?

This announcement makes a fair amount of sense. It fits in with the difficulties & delays in creating IMS-capable handsets, and the trend I noted the other day of fixed-VoIP being a lead IMS app, even for mobile operators, and often implemented on PCs.

Laptops are clearly even better than desktop PCs as IMS end-points from an operator standpoint - they have a reasonable chance of being "mobile" or at least nomadic, and they roam from time to time. (I'm writing this in an unseasonably cold & snowy Dallas). And PCs are much easier for developers to write applications on, they've got the horsepower & battery to do VoIP and multimedia, they're inherently multi-tasking and best of all, they're not usually subsidised.

Now, the announcement also talks about embedding HSPA in laptops. I've said before that I'm unconvinced about cellular embedded in PCs, for various reasons, and I still believe that it will not become widespread - maybe 10-20% of laptops shipped in a few years' time seems a reasonable range. On the other hand, a PC has a bit more room for a decent-sized antenna, so perhaps HSPA reception might improve vs a phone - although as with all 3G devices you're at the mercy of patchy indoor coverage.

On the other hand, unlike phones, all HSPA laptops will have both WiFi and a wired ethernet socket as alternative accesses, so the chance of getting some form of low-latency high-bandwidth connection for the IMS client is much higher, especially as once again the nature of the OS and the device makes it easier to write multi-mode connection software clients. They also don't tend to be used while actually walking around, so you can forget about all the overhyped distractions of "seamless handover".