AT&T proposes deploying LTE in WCS 2.3 GHz band

[iansltx]

Good to know. I wonder who would be willing to buy those AWS licenses, if AT&T decides not to deploy anything on them. We certainly don't need another SpectrumCo...

 

Which brings up another potential issue. If AT&T decides not to use AWS...ever...it has less spectrum to work with overall, so that it might be less amenable to a PCS-for-WCS swap than otherwise, even if the MHz-pop count is in their favor. Of course, that somewhat assumes that AT&T will want to drop LTE onto PCS now that AWS doesn't look so hot for them, but then they have issues if they decide not to ink a PCS LTE roaming agreement with Sprint, since both carriers would be deploying in PCS A-F. Not that that's stopped AT&T before (see the AT&T-championed lower-LTE-minus-band-A class versus the smaller-carrier-championed band class that includes lower-A), but PCS could be even more sketchy since the G band doesn't interfere with anything (other than AT&T's unwillingness to support a band that they won't be deploying anything on).

[ericdabbs]

Good to know. I wonder who would be willing to buy those AWS licenses, if AT&T decides not to deploy anything on them. We certainly don't need another SpectrumCo...

 

Which brings up another potential issue. If AT&T decides not to use AWS...ever...it has less spectrum to work with overall, so that it might be less amenable to a PCS-for-WCS swap than otherwise, even if the MHz-pop count is in their favor. Of course, that somewhat assumes that AT&T will want to drop LTE onto PCS now that AWS doesn't look so hot for them, but then they have issues if they decide not to ink a PCS LTE roaming agreement with Sprint, since both carriers would be deploying in PCS A-F. Not that that's stopped AT&T before (see the AT&T-championed lower-LTE-minus-band-A class versus the smaller-carrier-championed band class that includes lower-A), but PCS could be even more sketchy since the G band doesn't interfere with anything (other than AT&T's unwillingness to support a band that they won't be deploying anything on).

 

Well to be fair, I believe that AT&T still has a ton of spectrum to work with. You are ignoring the fact that they still have the 850 MHz cellular band of 25 MHz AND also have PCS spectrum which AT&T already says that they plan on deploying LTE in the future. I understand that GSM/HSPA+ is currently deployed in these bands but lets be honest at some point AT&T and the rest of the carriers will have to go all in on LTE once VoLTE becomes standard.

 

Its not like AT&T needs to have 2-3 LTE carriers at this point to support the huge capacity when they are still trying to fill up one LTE carrier. Even Verizon which has had LTE deployed for about 1.5 years the 10x10 MHz LTE carrier is still not full. Not to mention that AT&T hasn't talked about using further techniques like wifi offloading, cell splitting, small cells (picocells and femtocells) to deal with extending capacity on a LTE carrier like all carriers should be thinking of doing. Sprint has made that known about deploying small cells and wifi offloading to deal with capacity.

[iansltx]

Well to be fair, I believe that AT&T still has a ton of spectrum to work with. You are ignoring the fact that they still have the 850 MHz cellular band of 25 MHz AND also have PCS spectrum which AT&T already says that they plan on deploying LTE in the future. I understand that GSM/HSPA+ is currently deployed in these bands but lets be honest at some point AT&T and the rest of the carriers will have to go all in on LTE once VoLTE becomes standard.

 

Where did AT&T say they were going to start using PCS to deploy LTE?

 

Also, AT&T will have to keep 10MHz or so of its spectrum reserved for GSM + HSPA for the foreseeable future, because they've still got a TON of GSM-only handsets out there, and even more non-LTE HSPA phones. This is the bare minimum: one HSPA channel with no guard bands (3.84 MHz x 2), plus five 200KHz-wide (on Tx/Rx) GSM channels, which would only support 80 simultaneous conversations per cell with no data and all using the crappy half-rate codec (I think AT&T uses that anyway, meh). In order to maintain service coverage, AT&T would probably want to keep both GSM and HSPA alive on both 850 and 1900 bands, particularly considering just how little bandwidth the GSM side would provide, so that's 20MHz of spectrum that they can't use.

 

Its not like AT&T needs to have 2-3 LTE carriers at this point to support the huge capacity when they are still trying to fill up one LTE carrier. Even Verizon which has had LTE deployed for about 1.5 years the 10x10 MHz LTE carrier is still not full. Not to mention that AT&T hasn't talked about using further techniques like wifi offloading, cell splitting, small cells (picocells and femtocells) to deal with extending capacity on a LTE carrier like all carriers should be thinking of doing. Sprint has made that known about deploying small cells and wifi offloading to deal with capacity.

 

AT&T will have to use 5x5 LTE in a lot of places, which means its channels will fill quickly. Verizon is already saying that they want/need more spectrum than the 10x10 system they have, and they need it next year in some markets. That's what happens when you get spectrum that propagates way out there...you either have to turn down transmit power to add more capacity, or get an interference prone mess (that's the plus of Sprint deploying its primary LTE on PCS).

 

Circling back, GSM carriers are at a bit of a disadvantage on the LTE spectrum front, because subscribers have the expectation that the GSM-only phone they bought three months ago will still work three years from now, even as AT&T tries to push GSM spectrum to HSPA or LTE. Decreasing available bandwidth for HSPA will be even harder, though there will be a bigger windfall for doing so, since each HSPA carrier is 5MHz wide. The issue there si that you need to get the majority of your subscriber base onto LTE before you can start refarming HSPA, and when your LTE phones are expensive that just isn't happening quickly.

 

The nice thing about CDMA is there's only one technology that you have to support, and its channels aren't terribly wide (1.25MHz in each direction...wider than 200KHz GSM but much narrower than WCDMA/HSPA). Any phone made in the past ten or so years can use this network, though newer phones (1xRTT, EvDO, 1xA) are more efficient at using the airwaves. All that said, CDMA carriers will need to keep two of these carriers alive (one for voice, one for EvDO data) for the foreseeable future, using up 5MHz of capacity per band they decide to keep alive (in Verizon's case they would want to keep CLR online in markets where they have it...everyone else can just stick with PCS). But hey, 5-10MHz is a heck of a lot better than 20MHz when it comes to "un-refarmable" spectrum.

[ericdabbs]

Where did AT&T say they were going to start using PCS to deploy LTE?

 

My fault...=( I was mistaking AT&T with Verizon. Verizon did say they plan to deploy LTE on its PCS spectrum some time in the future.

http://www.fiercewir...mand/2012-06-07

 

I did however remember that AT&T was beginning to refarm its 2G spectrum in the PCS band in NYC to bolster their HSPA+ network. Either way I don't think AT&T should get a pass to obtain more spectrum just because they use HSPA+ which is more bandwidth intensive. The truth is that the majority carriers worldwide are moving to some form of LTE whether its FDD or TDD. If AT&T decides that it is in their best interest to support a full HSPA+ and LTE network then they must deal with the consequences of not having enough spectrum to support both. I believe that is the major reason why Sprint decided to forgo EVDO Rev B since it was just too bandwidth intensive when the obvious trend is towards LTE. Either way AT&T needs to start pushing hard towards LTE phones only.

[iansltx]

My fault...=( I was mistaking AT&T with Verizon. Verizon did say they plan to deploy LTE on its PCS spectrum some time in the future. I did however remember that AT&T was beginning to refarm its 2G spectrum in the PCS band in NYC to bolster their HSPA+ network.

 

http://www.fiercewir...mand/2012-06-07

 

Thanks for the link. It sounds like Verizon doesn't think it will have much LTE in PCS...almost undoubtedly less than Sprint...thanks to all those CDMA customers the company has. They also aren't in any sort of a hurry to deploy LTE there...2015 is 2.5 years form now! By that time, T-Mobile will probably have built its entire LTE network in AWS and refarmed HSPA+ to a PCS-heavy combo of AWS/PCS (where spectrum allows), and Sprint will have its entire network overlaid with LTE (in PCS and SMR). Heck, everyone will probably be using LTE-A, by then!

 

Of course, if VZW gets SpectrumCo goods, my bet is LTE on AWS will be up well before 2015. Cell spacing is close enough to PCS, and VZ can plop down a 10x10 carrier wherever it needs to that way, for 40MHz of LTE in areas that need it. That's a lot of capacity. Almost as much as Clearwire could provide in a hot zone...oh wait, Clear can do more

[bigsnake49]

Thank you, ian for doing all the legwork on the WCS for PCS spectrum swaps. I think it is in the best interest of both carriers to do the swap. I think the FCC will impose other conditions other than the sale of the 700Mhz A&B blocks. I think that they will also force them to sell their own AWS spectrum holdings concentrated in the eastern half of the country, at least in the large metro areas and Florida. Now Sprint also wants them to impose nondiscriminatory WiFi roaming on the cable cos rapidly expanding WiFi networks as well as backhaul:

 

http://www.lightreading.com/document.asp?doc_id=222163&site=lr_cable&

[iansltx]

Forcing Verizon to sell its own AWS holdings, at least in areas where they don't overlap with SPectrumCo, seems like a stupid idea. What Verizon is trying to do is get as close to nationwide on their AWS license (at 20MHz or so) as possible so they have economies of scale when building out their network. They can't be faulted for this...Sprint's got it with the G Block and ESMR, and Verizon's doing quite well with upper C block.

 

OTOH the conditions that Sprint wants imposed on Verizon with the SpectrumCo purchase are actually pretty reasonable.

 

The first big condition in the LR article is that Sprint (or anyone else) is able to WiFi offload to CableWifi at a cost that is reasonable on the wholesale side, and with an experience that's easy for customers. Put a different way, Sprint wants to make sure that cable companies don't give Verizon what amounts to a free WiFi-offloading ride that they would normally reserve only for other cable company customers...and leave any other carrier hanging when that carrier asks for access to the network. With everyone talking about spectrum and efficient use thereof, this seems like a perfectly reasonable request, since these WiFi networks have tons of capacity behind them and would make for an excellend offloading experience in areas (like NYC) that do actually need it.

 

Sprint's backhaul request is also reasonable, though I'm not quite sure how much they're asking for there. The gist I'm getting is that SPrint wants to be able to get comparable rates on backhaul to Verizon or anyone else, given of course quantity discounts, comparable speeds and similar endpoints. Again, the temptation here for the cablecos is to bend over backwards to give Verizon a good deal, since Verizon just bought a few billion in spectrum from them, and the money they save VZW on backhaul would probably come right back to them n the form of either lower wholesale rates on bundled services or better customer experiences due to a quicker LTE rollout. Not that these are bad things, but giving VZW a sweetheart deal on backhaul while inflating prices to Sprint, T-Mobile, etc. (perhaps to make up for lower VZW pricing) puts Verizon at an unfair advantage in a time where they're the most expensive wireless carrier anyway.

 

As a VZW customer paying $20-$50 per month for LTE iPad access, I understand why VZW and cable companies might want to work out something advantageous for both parties, but there comes a point (like with the AT&T&T failed merger) where antitrust is the bigger issue. I certainly don't want to be stuck paying $8-$20 per GB on this iPad for the next five years, and those rates are just as much, if not more so, about competition as they are about Verizon's own costs. I love me some Verizon LTE coverage (300+ markets now!) but if they are in cahoots with cable companies to make sure that they're the only provider that can get the pieces needed to ensure a fast LTE rollout, that's just a little too fishy.

[bigsnake49]

Forcing Verizon to sell its own AWS holdings, at least in areas where they don't overlap with SPectrumCo, seems like a stupid idea. What Verizon is trying to do is get as close to nationwide on their AWS license (at 20MHz or so) as possible so they have economies of scale when building out their network. They can't be faulted for this...Sprint's got it with the G Block and ESMR, and Verizon's doing quite well with upper C block.

 

 

They will have 20MHz (10+10) of near nationwide AWS if they're allowed to purchase cablecos spectrum. They will not have nationwide 20+20 AWS even if they're allowed to keep their own. They have nothing on the Western US. Actually, pre failed merger attempt, AT&T's AWS holdings were a perfect compliment to Verizon's.

[WiWavelength]

Actually, pre failed merger attempt, AT&T's AWS holdings were a perfect compliment to Verizon's.

 

Not exactly. Before AT&T partitioned and disaggregated selected pieces to T-Mobile, AT&T held the AWS D block 10 MHz license for the Central REA and the AWS E block 10 MHz license for the West REA. VZW holds the AWS F block 20 MHz license for the Northeast, Southeast, Great Lakes, and Mississippi Valley REAs.

 

So, yes, the licensed areas are perfect complements. But the licenses themselves are not.

 

AJ

[WiWavelength]

As a side note, Sprint owns WCS in other areas than the DFW MEA. They have 20MHz in Charlotte-Greensboro-Raleigh (MEA007), Atlanta (008), Tampa/Orlando (010), Louisville-Lexington (023), Nashville (025) and New Orleans/Baton Rouge (027), and 10MHz in Birmingham (024), Memphis (026) and Jacksonville (009). If anyone is intensely curious about how many people those licenses cover, I could spend some more time running numbers, or you can go to the FCC website and do the addition yourself.

 

Be careful with your data above. Most of it is partly incorrect (or, at least, could be misconstrued). BellSouth (i.e. AT&T), not Sprint is largely/entirely the WCS A block 10 MHz and B block 10 MHz license holder for the MEAs you list. Sprint holds only partitioned and/or disaggregated portions of those licenses, primarily outside of the major metros.

 

AJ

[iansltx]

Be careful with your data above. Most of it is partly incorrect (or, at least, could be misconstrued). BellSouth (i.e. AT&T), not Sprint is largely/entirely the WCS A block 10 MHz and B block 10 MHz license holder for the MEAs you list. Sprint holds only partitioned and/or disaggregated portions of those licenses, primarily outside of the major metros.

 

AJ

 

Hence my disclaimer about needing to go through it I thought I had grabbed only paired spectrum when I looked those up though. Maybe I didn't...

[bigsnake49]

Not exactly. Before AT&T partitioned and disaggregated selected pieces to T-Mobile, AT&T held the AWS D block 10 MHz license for the Central REA and the AWS E block 10 MHz license for the West REA. VZW holds the AWS F block 20 MHz license for the Northeast, Southeast, Great Lakes, and Mississippi Valley REAs.

 

So, yes, the licensed areas are perfect complements. But the licenses themselves are not.

 

AJ

 

Which makes my point above even more strongly. I don't think that the FCC will allow them to have 40MHz in the areas you listed. Maybe 30?

[iansltx]

Which makes my point above even more strongly. I don't think that the FCC will allow them to have 40MHz in the areas you listed. Maybe 30?

 

Maybe, maybe not. I think T-Mobile has 30MHz in some areas right now, and could hit 40MHz with the AT&T divestitures.

 

Remember, the AT&T spectrum package was part of the "breakup fee", and was not imposed by the government.

[WiWavelength]

Which makes my point above even more strongly. I don't think that the FCC will allow them to have 40MHz in the areas you listed. Maybe 30?

 

Yes, from the beginning, I have said that if the VZW-SpectrumCo-Cox transaction has any chance of FCC approval, then VZW will likely have to divest the AWS F block 20 MHz licenses that it has been squatting on for the eastern half of the US for the past six years. Basically, VZW would give up its current 20 MHz of AWS east of the Mississippi but get in return 20 MHz AWS for nearly all major markets nationwide (absent St. Louis, Cincinnati, Buffalo).

 

https://twitter.com/WiWavelength/status/173114138468425728/photo/1

https://twitter.com/WiWavelength/status/173114407734362112/photo/1

 

AJ

[WiWavelength]

Maybe, maybe not. I think T-Mobile has 30MHz in some areas right now, and could hit 40MHz with the AT&T divestitures.

 

I would not consider T-Mobile an apt comparison. Unlike VZW, T-Mobile does not have any Cellular 850 MHz spectrum (outside of Myrtle Beach, SC) nor any Upper/Lower 700 MHz spectrum. Not to mention, T-Mobile does not have the largest national market share.

 

VZW presents definite market and supply concentration concerns. T-Mobile does not.

 

AJ

[lynyrd65]

 

I would not consider T-Mobile an apt comparison. Unlike VZW, T-Mobile does not have any Cellular 850 MHz spectrum (outside of Myrtle Beach, SC) nor any Upper/Lower 700 MHz spectrum. Not to mention, T-Mobile does not have the largest national market share.

 

VZW presents definite market and supply concentration concerns. T-Mobile does not.

 

AJ

 

Do they actually use that spectrum there?

 

Sent from my Nexus S 4G using Tapatalk 2

[WiWavelength]

Do they actually use that spectrum there?

 

Lynyrd, I am not sure I follow your question. Are you asking about T-Mobile in Myrtle Beach?

 

AJ

[lynyrd65]

yeah, since they only have the one license there (it's actually one of the screwiest areas of the country spectrum wise) do they bother?

[WiWavelength]

yeah, since they only have the one license there (it's actually one of the screwiest areas of the country spectrum wise) do they bother?

 

Let me bring everyone up to speed with some background info. In all 50 states, T-Mobile has just one Cellular 850 MHz license for a single CMA in South Carolina because of its acquisition five years ago of Triton PCS dba SunCom. T-Mobile had essentially no native presence in the Carolinas, so it bought Triton, which had operated as the AT&TWS affiliate in the Carolinas up until the Cingular-AT&TWS merger in 2004. Triton had an almost exclusively PCS 1900 network, except for the one Cellular 850 MHz in the Myrtle Beach area.

 

Now, Cellular 850 MHz licenses have geographic area based construction requirements and currently utilize a site based coverage model (though the FCC is likely going to change that licensing scheme in the next year). Basically, a Cellular 850 MHz licensee must provide and maintain coverage to its established Cellular Geographic Service Area (CGSA). Otherwise, the licensee forfeits the right to any unserved area.

 

So, yes, Triton PCS had already constructed its Cellular 850 MHz market in South Carolina, and T-Mobile necessarily continues to operate it.

 

AJ

[WiWavelength]

Here is a screen cap of T-Mobile's coverage in northeastern South Carolina. I suspect that most will be able to discern the larger Cellular 850 MHz cells from the smaller PCS 1900 MHz only cells.

 

 

AJ

[iansltx]

Ah, SunCom.

 

I used to have a T-Mobile prepaid phone (unlocked Cingular Nokia 6010). When I was in NC, pre-Suncom-purchase, I could use my choice of AT&T or SunCom, since T-Mobile had no native network in the area. it was kinda cool.