x
Optical/IP

LTE Femtos in the Slow Lane

There's not much point in a Long-Term Evolution (LTE) Femtocell unless a user has a fast enough wired connection in the home to support it.

This was brought home to Unstrung when we talked with Sanjeev Verma, a founder of Airvana Inc. and VP of femtocell business and corporate development, in Las Vegas recently.

Verma's reasoning is fairly simple. All femtocells use a wired broadband connection in conjunction with the onboard cellular radio to provide better cellphone connectivity within a user's house. Cable and DSL connections provide average download speeds of 1.5 Mbit/s worldwide, although speeds in the U.S. are now up over 3 Mbit/s in many areas.

We don't yet know what average download speeds LTE will offer since there isn't a commercial network in the world that actually uses the technology. Operator and vendor tests suggest, however, that upper limits of 50 Mbit/s may be possible. Motorola Inc. (NYSE: MOT) was certainly getting 8-Mbit/s download speeds on an LTE drive at the recent CTIA show.

Forthcoming 3G femtocells, meanwhile, will offer download speeds of 1.4 Mbit/s, and more when High Speed Downlink Packet Plus (HSDPA+) services come in. Verma's point is that 3G will be good enough in this case, and the speed of LTE will likely be throttled by the user's cable or DSL connection.

"It doesn't matter if you have the fastest car on the freeway," says Verma, by way of an analogy. "If you're stuck behind someone doing 20 miles per hour, that's the speed you're going."

LTE deployments -- at least in the U.S. -- could also sweep away another argument for femtocells in the home, that the tiny home base stations improve call coverage. This is because the 700MHz spectrum that AT&T Inc. (NYSE: T) and Verizon Wireless plan to deploy LTE in is said to offer much better signal penetration in buildings.

For its part, Verizon Wireless is pushing ahead with 2G and 3G femtocells, while AT&T Inc. (NYSE: T) is expected to launch its 3G "microcell" sometime soon. Neither, however, have any announced plans for "4G" femtocells. (See MWC Preview: Femtocells Get Real.)

— Dan Jones, Site Editor, Unstrung

lrmobile_atiller 12/5/2012 | 4:07:09 PM
re: LTE Femtos in the Slow Lane

I tend to agree that the appeal of LTE femtocells will be somewhat limited unless residential broadband speeds can keep up (although there are some places where this is already the case).  However, if the femtocell is used as a means of connecting your mobile device into your home network, then backhaul is not required and the full speed of the LTE air interface can be used.  For example, UPnP over LTE could be used to access your home music and video collections from the mobile device.  To extend the freeway analogy, the home network is the equivalent of having a private racing circuit.

paolo.franzoi 12/5/2012 | 4:07:06 PM
re: LTE Femtos in the Slow Lane

Wow, I am confused at least from a US standpoint.


FiOS, U-Verse, and major Cable Operators support very high bitrate downstream connections.  Most of the Tier2 DSL operatiors are all working on 10Mb/s DSL offerings. 


Seems like this is an issue for somewhere else?


seven


 

joset01 12/5/2012 | 4:07:06 PM
re: LTE Femtos in the Slow Lane

Well, for one, we don't cater solely to a US audience. Also, the 3Mbit/s plus figure quoted was the current average for the US in March 09, despite faster connections coming down the pipe. FIOS and U-Verse have like -- what -- a million subs each?


DJ


 


 

paolo.franzoi 12/5/2012 | 4:07:05 PM
re: LTE Femtos in the Slow Lane

 


They each have more Dan.


And all the cable companies add to that.  They run 8Mb/s minimum.  So basically the entire US is covered by an 8Mb/s service.  That is not 100% true, but is true for all the metro areas that will have 4G.


 


seven


 

El Rupester 12/5/2012 | 4:07:03 PM
re: LTE Femtos in the Slow Lane

Indeed. He is obviously right: why should the inbuilding wireless be any faster than the access to the building? 


Which is of course why nobody bought 802.11g WiFi (54Mbps) and why 802.11n (108Mbps) has died - no-one wants faster than ADSL. 


.11b at 11Mbps is plenty fast enough for everyone isn't it?


</sarcasm>


Besides that fairly compelling precedent, there are other factors:


"we are not just for US market" - then you should know that many countries are already w-a-a-y faster broadband than the US rates you quote.


Japan & Korea 100Mbps to the home is standard.  In Europe, 20Mbps cable is widely available.  Even without fibre, DOCSIS3 is  50-100Mbps and will be universal in US before LTE reaches mass volume.


Fibre is happening: you say "only a million subs each" - how long will it be till Verizon has a million LTE subs...!


 


But thirdly, and most importantly, there is a fundamental misunderstanding.


Femtocells are not just residential.


Enterprise, metro, hot-spot, all can use the technology (cheap basestation, plug&play deployment), and will all can have faster backhaul than the mass market.


Indeed, given the fact that LTE only delivers real benefit at short range (for most carrier at 2GHz), then it is essential and predictable that LTE will primarily be delivered by small cells.


Domestic LTE femtocells will happen when LTE becomes mainstream, but LTE femtocells will become widespread earlier than that as part of many carriers deployments well before mass-adoption.


 


 

mobileinsider 12/5/2012 | 4:06:59 PM
re: LTE Femtos in the Slow Lane

I agree with my fellow posters.


LTE Femto is about 'when' not 'if' for US and Europe. But, a likely time-frame for enterprise and consumer LTE/HSPA combo Femto AP is closer to 2011-12 when spectrum constraints meet technology readiness. LTE only Femto?  More like 2014-15. (my 2 cents)


Twitter/mobileinsider

4Gconsultant 12/5/2012 | 4:06:56 PM
re: LTE Femtos in the Slow Lane

I feel, in next few years LTE will be adopted as main technology for mobile networks and there will be flood of handheld devices, mobile with LTE. This is is going to tigger the requirement for LTE femtos. Backhaul bandwidth should not be an issue, since a mobile user subcription plan will never be exceeding the broadband speed. Yes, one need to think about backhaul problem for enterprise LTE femtos.

vsomanv 12/5/2012 | 4:06:56 PM
re: LTE Femtos in the Slow Lane

- some researchers would vie for LTE, and some would for WiMAX and some would say 4G is not required for years to come


- LTE should be ubiquitous, and that is basically got to do with Voice. Voice over 3G and WiMAX is a big distance dream. You could see a patchwork of VoIP over 3G and WiMAX here and there. But does no good to the equivalent of GSM and the likes


- Device is the king. The technology bringing in more number of handhelds win the game


- Femto is all about penetration indoor and less to do with other stuff. 450 MHz, 700 MHz, 800 MHz and 900 MHz folks would require less or no Femtos. > 1 GHz folks would require some or more.


- enterprise Femtos or Super Femtos or the lower end Picos would not have much of Backhaul issues (They already have a WAN Link in order)


- WiMAX would remain a Wireless DSL all its life.


- Free spaces in 700 MHz could create some uproar


- Ofcom has suggested to free some bands in 800 MHz, good news for LTE

IPobserver 12/5/2012 | 4:06:54 PM
re: LTE Femtos in the Slow Lane

If femtocells make sense for 3G, surely the concept applies to LTE. Maybe it’s a question of timing. 

 Lots of points to think about. Off the top of my head:

 * What handsets apps are likely to require more than, say, 3 Mbit/s given the screen size, processor, implications for battery life, etc?

* What apps would require a “private race track” ? – data back-ups, adding videos/music to phone, I guess

* What sort of channel widths are you talking about for LTE femto? Is it TDD or FDD? Which frequency?

* Has anyone got the stomach to invest in LTE femto development right now? – Not sure. Existing players need to see 3G pan out first

* Are there inherent advantages to the LTE radio over 3G and WiFi? (e.g. for battery life, active/idle transitions, latency, QOS, etc)

* How will the more generic “small cell” market evolve?

* Loads more stuff I can’t think of right now

VZW CTO talked about LTE Femtos at last year’s C-Scape. See here.

vsomanv 12/5/2012 | 4:06:46 PM
re: LTE Femtos in the Slow Lane Quite coincidental, Verizon has released their specifications for Open Devices on LTE. They have indicated quitely, but strongly, that it will be FDD over LTE.

Not sure whether Femto would pick up though in this scenario. Goes on to assert that LTE Femtos are DEFINITELY in the Slow Lane.
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE