Nokia Promises EDGEy 2003

Nokia Corp. (NYSE: NOK) claims it will deliver EDGE-capable mobile devices to European GSM operators in the second half of 2003.

Rene Svendsen-Tune (no, he didn't sing us a song!), senior VP for marketing and sales at Nokia's IP Mobility Networks division, told Unstrung at Nokia's Mobile Internet Conference in Munich that "EDGE terminals for European frequencies [900 MHz and 1,800 MHz] will be available in the second half of 2003," though he would not elaborate on volumes.

He also declined to name any of the operators that have allegedly taken delivery of EDGE (enhanced data for GSM evolution) equipment, but noted that operators that do not have a UMTS license are likely to be a key target market.

"EDGE is an obvious next step for those operators that do not have a 3G license. We are convinced that those GSM operators that do not have 3G spectrum will go down that route. We also believe that most GSM operators will leverage their [2G] spectrum to deploy EDGE at some stage, but there is not a strong momentum in Europe just now."

EDGE is an enhancement to the GSM and TDMA wireless communications systems that increases data throughput up to 384 kbit/s. EDGE uses the same basic network structure as existing 2G technologies. Nokia has already made some loose announcements about its EDGE position and plans (see Nokia Gets EDGEy). For those operators without a 3G license, EDGE would at least offer an opportunity for higher data rates than will be possible with GPRS, while, in many cases, not having to replace base stations. Both Nokia and LM Ericsson (Nasdaq: ERICD) have built upgrade opportunities into their GSM network equipment for years, and they market EDGE technology as a simple and cheap upgrade. In addition, as part of the GSM evolution cycle (GSM-GPRS-EDGE-Wideband-CDMA), it would allow these 3G-less carriers: applications continuity; enhancements enabled by increased data rates; and the possibility of device support from existing GSM vendors.

Executives at Ericsson, Nokia's main GSM infrastructure rival, agree that EDGE will be deployed first by the 3G-less operators, and ultimately by most GSM carriers. "Why wouldn't an operator want to triple the IP capacity offered by GPRS?" said Mikael Halén, the Swedish vendor's director of W-CDMA marketing during a recent telephone interview with Unstrung. "GSM operators want to preserve the investment they have already made in their networks."

Which all sounds rather promising for the takeup of EDGE, seeing as how no additional (expensive) licenses or spectrum are required. Nokia's Svendsen-Tune is wise to this and, without prompting, proclaims W-CDMA as the ultimate capability for GSM operators and startup greenfield carriers alike. Wideband-CDMA increases data rates in GSM systems by using the CDMA air interface instead of TDMA; it is part of the universal mobile telecommunications specification (UMTS).

"W-CDMA still allows the most cost-effective solution and the best use of available capacity," he proclaims, knowing of course that the 3G license conditions require network infrastructure above and beyond the capabilities of EDGE. Knowing that they need to pump their cash and effort into UMTS/W-CDMA to meet their regulatory requirements, European operators with 3G licenses have no plans at present to deploy EDGE, according to IDC senior research analyst Paolo Pescatore.

While Europe looks at present to be less of an opportunity, the market is already developing in North America, while Asia is showing signs of interest, says Svendsen-Tune. "There is clear progression in the U.S. where WCDMA frequencies [2.5 GHz] are not available yet. AT&T Wireless Services Inc. [NYSE: AWE], Cingular Wireless, and T-Mobile USA have all announced plans and begun work on their networks. EDGE terminals for North America's GSM frequencies [850 MHz and 1,900 MHz] will be available in the first half of 2003," he says, which indicates a slight slip in timescales for handset delivery, as Nokia executives said during the company's recent financial statements that they would be available before the end of 2002.

Asia is less developed, but some operators should deploy EDGE in 2003, says the Nokia man, but that this would be outside the territories already well developed in high-capacity network systems (i.e., Hong Kong, Japan, Korea, and Taiwan).

On more general matters, Svendsen-Tune says Nokia is delivering WLAN equipment and systems to mobile operators for WLAN/WAN integration on all continents -- though you can count the number of customers at present on the fingers of "several hands" -- and that this is its own manufactured equipment based on the Institute of Electrical and Electronics Engineers Inc. (IEEE)'s 802.11 standards. Its channel via systems partner IBM Corp. (NYSE: IBM) is very important here, he says.

In the W-CDMA network equipment market, "Nokia has said we will take 35 percent of the market, and we are already very close to that." How close? "More than 30 percent, measured in actual sales booked."

He also claims that Nokia has a "very strong market position in GSM networks in China," but declines to say what that equates to in market share, simply stating that Nokia has a global share of the GSM equipment market "in the high 20s percent."

— Ray Le Maistre, European Editor, Unstrung
spc_King 12/4/2012 | 9:19:08 PM
re: Nokia Promises EDGEy 2003 The final nit-pick on EDGE:

I agree with both 'standardsarefun' and 'wirelessundertaker's comments and will add: you can even add AMR to both regular GSM (GMSK modulation) as well as an EDGE variant AMR solution (using 8-PSK for voice as well as data (CSD/GPRS)).
I've seen test results indicating 50% datarate increases using EGPRS coding schemes over and beyond CS-2 GPRS for the same cell radius that AMR 12.2 would be used (that is probably as big or even bigger cell radius than in current cells for networks using FR/EFR codecs).
At 80% of the AMR 12.2 radius you get x2 data rates using EGPRS and at 20% of AMR12.2 cell radius you get the full x3 datarate increase.

AMR using 8-PSK is GERAN 3GPP release 5 stuff and some time away, though. However. AMR using GMSK has been standardized since ETSI'98.

I don't think 'berzerk' has any more valid excuses why operators shouldn't use EDGE, especially considering the recent press-release on Nokia EDGE handsets - Eat your hart out, Berzerk! :-)

WirelessUndertaker 12/4/2012 | 9:19:21 PM
re: Nokia Promises EDGEy 2003 > 1) EDGE is not something separate from GPRS it is simply a new set of channel codes (and a new modulation). A phone that "does" EDGE must also "do" GPRS

Actually this is not true. EDGE can run on both circuit switched (called E-CSD) and packet switched (called E-GPRS). In other words, the EDGE modulation and coding schemes can be applied to both circuit switched and packet switched GSM networks. In the circuit switched mode, you'll need high speed circuit switch data (HPCSD) to support EDGE rates.
standardsarefun 12/4/2012 | 9:19:31 PM
re: Nokia Promises EDGEy 2003 Some of the previous posts seem to have missed a few points:

1) EDGE is not something separate from GPRS it is simply a new set of channel codes (and a new modulation). A phone that "does" EDGE must also "do" GPRS and the vendor would have to be stupid to not also "do" GSM voice. Same thing applies to base stations and so you will not need a have "EDGE carriers" and "GSM/GPRS carriers" but would tend to have "old carriers" (GSM/GPRS only) and "new carriers" (GSM/GPRS/EDGE) and so the traffic engineering is not as hard

2) Cell edges would tend to use the "old" channel codes (long live CS1) with the middles of "EDGEd" cells using the "new" (EDGE) channel codes. This means that your cell engineering remains the same (no new cell sites to deploy EDGE except for normal cell splitting as the traffic builds up).

3) BUT the average (over time and over all users) bitrate will not be enormously bigger (my earlier posting suggested x2 only).

4) You could do better if you deployed EDGE only carriers (no voice or normal GPRS terminals would be assigned to them) but then you would either need smaller cells or a wider reuse pattern and I don't see an operator having the spectrum unless them deployed this in rural areas using their 3G licences
BT76 12/4/2012 | 9:21:14 PM
re: Nokia Promises EDGEy 2003 current Euro operator thinking is the following:

* I already have GPRS rolled out
* I HAVE to deploy UMTS (at least in the cities) to keep the license
* When higher speed data becomes a requirement, EDGE will be added outside the cities to complement UMTS

First-gen UMTS handsets will be UMTS only. The thought is to add EDGE to the UMTS terminal per the assumption above.

No current commercial talk of EDGE anywhere outside the US (and only there from what I've explained earlier).
spc_King 12/4/2012 | 9:21:28 PM
re: Nokia Promises EDGEy 2003 Berzerk wrote: "user would require 2 terminals)"

Not that I have seen an EDGE phone yet, but I doubt that you would need one phone for voice and one for EDGE data - I'm sure that would be combined!?

Someone with access to both WCDMA and EDGE basestation and handset cost estimates would need to set up a detailed model to assess which scenario is most cost effective, but my gut feeling is an EDGE mix would be a lot cheaper!

BT76 12/4/2012 | 9:21:51 PM
re: Nokia Promises EDGEy 2003 X-Eri, you're right about MOST of this. And sorry, no disrespect intended.

CS3 and CS4 would screw up the RF coverage/footprint drastically (currently CS3 and CS4 and not implemented in any hardware). Add that to the already decreased footprint (with respect to current GSM/GPRS) from a higher modulation scheme and there would be too many coverage holes to actually offer/advertise the service. Carriers will not (financially) be able to add the required sites to fill these holes and you can't sell new EDGE phones if there's not a ubiquitous EDGE service offering.

However, you're right with respect to EDGE complimenting UMTS long term: UMTS will be in the urban centers and EDGE may come along later in the rural areas. My point is that the migration path is NOT GPRS->EDGE->UMTS but GPRS->UMTS (because they have to to keep the licenses)->EDGE outside major cities (when high bandwidth service ubiquity is warranted--when there's $$$ to be made).

At this stage, 3G will be built because operators have ALREADY paid a boat load for the licenses/spectrum and the investors/shareholders (and the governments) won't allow them to stop halfway. Just think about it: there's still no business model for high speed wireless data but all they systems are being built anyway (and in this economy too).

In the US, CDMA EV-DO is now a finalized standard but no major carrier will deploy this commercially for at least another 2 years. Why? There's still no biz model for high speed (or even low) speed wireless data yet.

Finally, as operators HAVE to build UMTS and now have GPRS, there is no current reason to build anything EDGE (an in-between technology that originally was for GSM carriers who weren't going to get UMTS spectrum). Remember, EDGE is currently data-only (UMTS does both) so operators would have to supplement (first/currently) a GPRS phone, then an EDGE one (and just for data so the user would require 2 terminals) and finally a UMTS phone. Again, what do I as the user need these phones for? This just doesn't make sense and everyone will realize this as time goes by. However, there is talk (long term, not short) of a UMTS/EDGE combined phone for the scenario above.
spc_King 12/4/2012 | 9:22:13 PM
re: Nokia Promises EDGEy 2003 More than just one Boo-Boo imho :-)

Yes, I probably do think like a vendor...

However, the reason I see an opportunity for EDGE even in Europe is that I think WCDMA licencees can create a cheaper 3G system if they combine WCDMA (in hotspots) with EDGE (in more rural areas).

Seen in the light of having to build out 90-100% coverage in a country using WCDMA, I would think EDGE and CS3-4 basestations would seem a really good deal comparatively, especially since (as you point out) there is no demand for the fatter WCDMA pipe.

If operators get scared by the cost of EDGE or basestation upgrades then what the hell are they doing investing in 3G (be it CDMA2000 or WCDMA)??

EDGE can save GSM/WCDMA operators' bacon. Please show som more respect :-)

BT76 12/4/2012 | 9:22:16 PM
re: Nokia Promises EDGEy 2003 one 'boo boo' in the next to last paragraph of my previous post---

Sites DON'T go further with higher modulation...they go a shorter distance (which means you need MORE sites for equivalent coverage).
BT76 12/4/2012 | 9:22:16 PM
re: Nokia Promises EDGEy 2003 Hey Ray...here you go:

10% anytime soon (next 5 years). Who can say after that but we're already seeing comoditization in data in the U.S. Ultimately (and wireless carriers don't want to believe this), they will give 'all you can eat' data and be a pipe. There's no other way these HUGE UMTS pipes will EVER be filled. They will ultimately be another DSL provider with a wireless and speed premium over fixed DSL. Again, no other way to fill these HUGE pipes.

Its already clear in Europe that EDGE will come AFTER UMTS as UMTS will not be ubiquitious (too expensive to put it EVERYWHERE as its above 2 GHz and therefore A LOT of BTSs would be required). Handset manufactures don't have the money to support too many standards...it looks like GPRS/UMTS/1X-CDMA will be it.

If it wasn't for AT&T and Cingular (here in the US) needing a comeback to Wall Street for the fact that they have no high speed 1x->EV-DO->EV-DV roadmap, EDGE would not even be around. Remember, with EDGE you have higher modulation, that means sites go further (no free lunch in RF). That means more sites. And unlike GSM, EDGE does NOT carry voice traffic. The GSM system is going to have to be MIGHTY loaded before a separate channel is allocated in the ENTIRE network for data only. How about just 1 EDGE TRX @ US$10,000/trx X 3 sectors/site X 1000 sites in a metro area = US$30 million (per big area) + backhaul + handset subsidies.

At what point does data warrant this (and, by the way, this is JUST 1 trx=8 time slots per sector)?
spc_rayella 12/4/2012 | 9:22:17 PM
re: Nokia Promises EDGEy 2003 Thanks for the post. I want to question a few of your assumptions.

Data will NEVER be more than 10% of voice ARPU? Is there a rational explanation for this? Do you mean globally EVER, or anywhere soon?

And regards 'No need to move from GPRS to EDGE and THEN to UMTS.' What about the carriers that do not have a UMTS license but have an existing GSM network that can be upgraded to EDGE?

I am not saying it isn't too much more than avendor wish,as the potential market might actually be too small for any meaningful econoimies of scale, but I am not sure it's as cut and dried as that.

Unstrung Ray
BT76 12/4/2012 | 9:22:21 PM
re: Nokia Promises EDGEy 2003 x Eri....you're missing a couple important points:

* No plans ever for CS3 or CS4. System will ALWAYS be RF planed for voice (as the cash cow). Anything beyond CS1 (normal GSM error coding) will never happen (on a grand scale as its too expensive for the operator). Remember, data will NEVER be more than 10% of voice ARPU.

* Combining GSM time slots above 3 requires the phone to have a duplexer as now the handset has to transmit and receive at the same time (that means more $$$, more battery, more loss, etc.)

If/When GPRS is successful, UMTS will be mature (in about 5 years). No need to move from GPRS to EDGE and THEN to UMTS. This is a vendor wish list...not a carrier's.
spc_King 12/4/2012 | 9:22:44 PM
re: Nokia Promises EDGEy 2003 EDGE only doubles? Hmm, I am pretty sure the modulation scheme offers 3-three times the bitrate, but one may need to discount it a little bit for highere sensistivity to weak signals etc.

If GPRS today gives you 20-30Kbps with CS2 then EDGE should give you 60-90kbps with the same coding scheme (i.e. same ratio of error correection bits to data bits).

The reason that is nowhere near the 384kbps theoritical max is because todays GPRS solutions only use half the amount of timeslots that are theoretically available in a given cell (i.e. 4 instead of 8). This is partly to avoid blocking a cell to voice calls and partly due to battery/heating issues.

However, both 'classic' GPRS and EDGE can increase performance over the 20-30 (60-90) of today once operators take delivery of basestations using CS3 and CS4 (and the new EDGE ones). These higher coding schemes demand better link conditions and means you either have to be lucky, have handsets with better antennae (e.g. Diversity) or more realistically operators need to install basestation HW/SW for network based antennae diversity (3db improvment is typical and gives a hell of a boost to the perfomance!)

There is hope yet for GPRS :-)

BT76 12/4/2012 | 9:22:57 PM
re: Nokia Promises EDGEy 2003 no one will really know until its deployed and optimized but here's some food for thought:

* EDGE modulation scheme doubles over GPRS (8PSK)

* EDGE is data only so the link efficiency goes up (no need to traffic control both voice and data)

Now for the opinion phase of this post: EDGE was a stillborn technology for carriers like AT&T/Cingular who had nothing to say to Wall Street to combat CDMA's roadmap. However, if we ever get to the stage where we need it, all of us in wireless data are winners.
standardsarefun 12/4/2012 | 9:23:08 PM
re: Nokia Promises EDGEy 2003 Does anyone have ideas about the real bit rates a real users would see in an EDGE powered cell when it (and its neighbours) are carrying real traffic?

The famous 384 kbit/s obviously is not realistic since this means all 8 slots dediciated to one user (how many slots will these Nokia EDGE phones support - my guess it 4 maximum) and you only get the maximum bitrate per slot when:
1) you are in the middle of the cell
2) the rest of the world is asleep (2 AM will be a great time to mobile web surf!) so you don't share the slot
3) nobody is in the middle of a neighbouring cell surfing and the interference is zero

Given all of that what do you think the real average increase will be? I guess is about +50% over classic GPRS.

P.S. Remember you need to consider layers 1, 2 and 3 plus the traffic and cell engineering or else you need about 15 real EDGE cells and about 1000 phones before you get a good estimate
futureisbright 12/4/2012 | 9:23:45 PM
re: Nokia Promises EDGEy 2003 The interesting thing about wireless development is that you don't need terminals until the infra is ready, and you don't need the infra until the terminals are ready. And you don't need either until the operators are ready. Interesting balance to keep with cash flow requirements!

NOK, ERICY, and SI have sweept much of the GSM market in NA based on their promises for EDGE development. What are we looking at now, 1H04 customer deployment?
spc_rayella 12/4/2012 | 9:24:01 PM
re: Nokia Promises EDGEy 2003 "EDGE terminals for North America's GSM frequencies [850 MHz and 1,900 MHz] will be available in the first half of 2003," he says, which indicates a slight slip in timescales for handset delivery, as Nokia executives said during the company's recent financial statements that they would be available before the end of 2002.

Then, I guess, it's a case of deciding just what availability means, in what numbers etc. The only people that really know are the h'set guys themselves, though I'm sure the U.S. operators will have some dates they're keen on too!

Unstrung Ray
futureisbright 12/4/2012 | 9:24:02 PM
re: Nokia Promises EDGEy 2003 when will the EDGE 800/1900 terminals become available?
incards 12/4/2012 | 9:24:04 PM
re: Nokia Promises EDGEy 2003 Unstrung Ray:My previous post said 2003 (Nokia EDGE terminal shipping) and should have said 2002. Make that an EDGE 850/1900 MHz terminal. - IC -
incards 12/4/2012 | 9:24:04 PM
re: Nokia Promises EDGEy 2003 Unstrung Ray: I appreciate the update on EDGE and am sorry to here the press wasn't in on the EDGE Update 2002 event. Nokia's Q3 CC slides suggested that they would ship an EDGE terminal in 2003 and if they do I suspect it will be for IOT and conformance testing. Nokia is hosting an analysts breakfast in Vegas on the 17th in conjunction with Fall COMDEX 2002 and I am wondering if they plan to do a product launch there or at their Dallas Strategy Update in Dallas on December 3, or wait for CeBIT and CTIA. - IC -
spc_rayella 12/4/2012 | 9:24:07 PM
re: Nokia Promises EDGEy 2003 I did ask Nokia if I could attend, but the EDGE update and all the sessions on the Tuesday and Wednesday of this week in Munich were for customers and partners only, about 1,000 of them. No prying press allowed :-)

So we had our presentation on Monday, the outcome of which you can see on the sie here. EDGE was not included in the main presentation, which focused on the new terminals and some new services for carriers. EDGE is something that I tried to tackle in a one-to-one discussion with a Nokia executive.

EDGE is, I think, an interesting development, but I do wonder whether it will offer the vendors affordable economy of scale in their production plans, especially in terminals. The out-of-factory and unsubsidized prices for EDGE handsets could be quite steep, I would guess. We'll see.

Unstrung Ray
incards 12/4/2012 | 9:24:15 PM
re: Nokia Promises EDGEy 2003 Unstrung Ray: Nice job at the Sonera event. I am hopeful that you are attending the 2002 EDGE Update at NMIC and will give us a piece on the demo's and the roadmap of EDGE terminals. - IC -
spc_rayella 12/4/2012 | 9:24:35 PM
re: Nokia Promises EDGEy 2003 Thanks for that. It was not Dan who quoted the wrong frequency but me. Now corrected.

memo to self - take more time over the figures...

Unstrung Ray
christer44 12/4/2012 | 9:24:43 PM
re: Nokia Promises EDGEy 2003 Dan wrote:
>"EDGE terminals for European frequencies [800 >MHz and 1,900 MHz] will be available in the >second half of 2003," though he would not >elaborate on volumes.

Last time I checked Europe used 1800 MHz and not 1900 for GSM...

You don't really expect to get away with such obvious typos without nit-pickers calling attention to it, do you? :-)

Sign In