I've been at VON in San Jose this week. This morning I moderated a panel session on dualmode WiFi/cellular handsets, and putting together the intro slides made me crystallise my thoughts on them a bit more clearly, as I seem to have contradictory optimistic/pessimistic thoughts about them.
In a nutshell, dualmode phones are:
a) Qualitatively more important than many people think
b) Quantitatively less important than many people think
Point (a) refers to the catalytic nature of WiFi-capable phones - stimulating new business models & service providers in the mobile space, teaching end-users about using phones to connect to non-service applications and devices (PBXs, PCs etc), driving diversity & maturity into the non-carrier channel for phones, and so forth. The inclusion of WiFi also points out the absurdity of paying a premium for "mobility" when you're not actually moving about.
One of the panellists this morning said that the average number of cell base stations used in the course of a call is just 1.06. In other words, less than 6% of calls make use of the hugely complex & expensive cell-to-cell handoff capability in the network; the remainder is in a single cell, and often in range of a single WiFi AP.
Point (b) is the counter to this. I keep seeing forecasts from my analyst peers about hundreds of millions of dualmode phones shipping in the near future. I think this is wrong - I honestly can't see the "attach rate" of WiFi going beyond 10%, maybe 15%, of phones. It's so much more than just adding a $5 chip - it's the software, the integration, the application & UI optimisation work.
And when you think in a bit more detail about the handset market, the addressable segment of phones for which WiFI makes sense is actually quite small. So, for example, about 60-70% of phones worldwide are sold to prepay customers, often sold independently of any service provider customisation. In my view, I see almost no justification of including WiFi in these devices - certainly, I'd bet almost no dualmode devices have shipped into the prepay market to date.
Then there's CDMA - about 20%-ish of handset shipments. There has been much less interest in CDMA/WiFi dualmode to date - no equivalent of UMA, and not much interest in VCC or proprietary SIP-based dualmode. Sure, this might change, but not immediately.
Then there's the simple issue that outside N America, Europe & a couple of Asian countries, there's simply not that much residential broadband, let alone WiFi-enabled APs or gateways.
And, lastly, there's so much other cool stuff you could put in phones - GPS, VGA screens, TV, motion sensors - that WiFi is competing against lots of other hardware & software for space on handset designers' priority lists.
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
6 comments:
I'd bet almost no dualmode devices have shipped into the prepay market to date.
Dean i am not a big flag waver for wifi phones but couldn't such a device actually be the ultimate in prepay?
Buy the phone with a skype client and voila... Even developing markets are getting public/muni wifi and hotspots. and if there is a real app - voice - ad supported maybe - why wouldn't chai/chaat shops in india, tea and dumpling shops in china want to put up an AP just as coffee shops in the NA/Asia/EMEA did?
Don't underestimate the CDMA market either. With the purchase of Airgo by Qualcomm, Qualcomm is quickly repositioning itself from a Wi-Fi hating chipset vendor to a 'Wi-Fi Enabler'.
I thought this comment very intresting. Dean, can you tell us who reported this? Do you agree with the statistic?
"One of the panellists this morning said that the average number of cell base stations used in the course of a call is just 1.06."
Bruce - I's see that opportunity as more one for single-mode WiFi rather than dual-mode. Also bear in mind Skype is also working with operators to put handset clients on phones which don't do VoIP, but dial into a fixed Skype gateway instead (eg 3 in the UK)
Anonymous - sure, I'm expecting some WiFi/CDMA eventually, but not in significant volumes for quite a while. There could be some CDMA-based VCC dualmode service launched at some point, maybe Sprint+cable operators for example. Not happening quickly, though.
Bailey - comment was from a representative now working for Truphone, who said he'd analysed 20 cellular networks for that data. Sounds eminently reasonable to me - if anything, I could have believed 1.03 or even lower.
I agree dual phone in CDMA market takes a while to happen, i think the issue is not from the terminal side, and my understanding is the CDMA network operators are still working on how to cooperate with those SIP servers and offer ultimate seamless handoff between wifi and cellular.
and i believe this dual phone has huge potential considering wifi has free spectrum, it exists almost everywhere in the city,thus end users could shrink their phone bills to some extent, and what else could beat that?
Hello Dean,
Great blog!
You said: "In other words, less than 6% of calls make use of the hugely complex & expensive cell-to-cell handoff capability in the network; the remainder is in a..."
I don't quite agree with the statement that the network is complex & expensive because handovers which are only required in a few cases :-) Here's why:
The same "hugely complex & expensive cell-to-cell handoff capability" also ensures that when you leave a place the phone stays connected to the network and is able to respond to new incoming calls. Sure, this cell reselection is initiated by the phone rather than by the network which controls the process during a call but from the network side the same datafill is required to let the mobile phone know about neighboring cells. Also the coverage must just be as optimized for cell reselection as for handover as otherwise the mobile phone would loose the network every now and then which would result in missed or delayed calls, SMS, eMail, etc.
Also, it should not be forgotten that even if no handovers are required for a phone call the network still controls the power level of the call which is part of the handover process to start with. Without that it would not be possible to have calls even without handovers.
So from my point of view, the handover mechanism is just a tiny little piece on top of functionality that has to be in place for every call anyway. Thus, I think networks could not be a lot easier or cheaper without handovers.
Cheers,
Martin
Post a Comment