How Exactly Will AT&T Remain Competitive?

[BuffaloTF]

Let's be fair now.

You forgot to mention EVDO was still a major upgrade because its a totally new all IP network..

So Verizon

1. Analog to CDMA
2. CDMA1X to EV
3. Combining Alltel
4. Centennial and Unicel which required GSM to CDMA
5. CDMA to LTE

How isn't it fair, exactly? 1X and EVDO are the same things, CDMA2000... EVDO is an enhancement for data on top of what 1x already did. If you want to be so technical as to split the two of them as these huge separate things simply because they had to add 1 new antenna to their towers and do a software upgrade - then to be fair, AT&T would look more like this:

AT&T:

1. Analog to D-AMPS.
2. D-AMPS to GSM.
3. GSM to GPRS
4. GPRS to EDGE
5. Combining Cingular and ATTWS.
6. GSM/EDGE to UMTS/W-CDMA voice.
7. UMTS to HSDPA 7.2
8. HSDPA to HSPA (which includes HSUPA in addition to the HSDPA)
9. HSPA 7.2 to HSPA+ 14
10. HSPA+ 14 to HSPA+ 21
11. HSPA+ 21 to LTE.

[zephxiii]

HSUPA came with the 3.6mbit upgrade. Before they went to 3.6mbit they were at 1.8mbit (I remember this era).

[GSMinCT]

Perfect.

If AT&T can give everyone in Verizon LTE areas at least 5-6mbps, that will be fine for a good while. Heck, Verizon is only promising 5-12 with LTE!

Verizon is certainly putting the pressure on AT&T, which is fantastic. Sorry, but AT&T sucks at network management, and even though it took Verizon's agressive LTE rollout to do it, I am glad they are finally giving users the speeds they need. As you stated, nobody can argue with saying EDGE has GOT TO GO! Verizon has no choice but to roll out LTE there, but AT&T can be fine with just a good, well built-out HSPA+ network in these rural areas. Give LTE to the masses, but give the rural areas something other than EDGE, and HSPA+ is a solid choice, assuming it is properly constructed.

Of course, at the same time, if AT&T goes out to these rural sites to hang HSPA+ panels, why not go ahead and hang LTE panels if spectrum permits? Who cares if they aren't active for a year or more! If AT&T were all HSPA+ today, Verizon's mid-2013 LTE completion would be far less important. I just want to know if AT&T can keep up with even HSPA+...

Yeah, HSPA+ is going to give AT&T a lot of breathing room on LTE, and they can even advertise that they have "two 4G networks" which is basically BS, but sounds good.

Yeah, AT&T sucks at network management in some places. Here in CT, they are top dog, and are 5x faster than Verizon, not even counting AT&T LTE (we don't have Verizon LTE yet).

EDGE has to go for marketing reasons. Personally, I'd be happy with AT&T fixing EDGE so that it consistently gets 150kbps and can stream Pandora on the highway, but I've given up on that, and seeing that they need HSPA+ from a marketing perspective, that's what I'm hoping for as a solution for broken EDGE.

I don't think AT&T can keep up with HSPA+ to be honest. If they had their whole network running at least HSPA with T1's, I would say there is hope, but considering they don't even have the air interface side of HSPA+ up on half their coverage area (although far less than half their towers), it's not looking too promising. Hopefully, this little fire that Verizon lit under their @$$ will cause them to speed up and get HSPA+ out there, and make it work properly. We also know that while it doesn't apply to my 14.4 iPhone, HSPA+ 21 devices can get the same 5-12 mbps speeds that Verizon promises on LTE if it is properly implemented.

Most rural sites, the only major upgrades they are doing to enable HSPA+ are replacing base stations and adding channel cards + adding fiber optic backhaul or enhanced ethernet over microwave, they typically don't have to have a tower crew come out and replace anything. We have HSPA+ running over Omnidirectional's here..

That would be great. If that's what they are doing, I wonder why they can't get it done faster, since they don't have to do any work to the tower itself. I was also wondering about that, since there is an area in Alaska that has 3G, but yet clearly still has Dobson Yagi's, and that must be the answer.

WTF no they do not need to skip HSPA since if they are installing LTE they are installing equipment that very easily supports HSPA. And given how huge the HSPA capable user base is it would be dumb as hell to skip it. You are better off going HSPA first then LTE later than anything.

Exactly, that would be just idiotic. I don't want to be stuck on EDGE while the guy next to me is pulling 50mbps on a Skyrocket.

[mush10]

The big question to me is whether Verizon's LTE network can stand on it's own. Meaning they will absolutely have zero need for people to drop to EVDO. I have had a Verizon LTE device since they have launched and my speeds have been getting progressively worse. In some ways it was a disservice to these companies to see speed tests coming through at crazy speeds when all the promised is 5-12. I am perfectly content if I end up with a consistent 5-12 on a network that can handle it even during the busiest times. Right now, I do have some doubts, especially with a LTE iPhone on the way. Is Verizon allocating enough LTE spectrum for a congested network. If not, AT&T's approach could be the winner.

This is where AT&T "could" separate themselves. As many early UTMS, HSDPA, HSPA adopters know there were plenty of times where forcing EDGE/GSM was your best bet for data or texting. If AT&T builds a massive LTE network with fallback to a consistent HSPA+ network they are in great shape. The problem with AT&T lies in the execution and timing. Even if I have LTE in major cities and know I am getting 4-10 on HSPA in rural areas, I would be all over AT&T, but I am not so sure that will happen.

A couple questions, does Verizon have enough 700 spectrum to effectively run a fully loaded LTE network? How does that compare to AT&T running LTE in 700 with ability to move more 850 and 1900 to HSPA+ and possibly LTE in the future? Will Verizon move all of there CDMA network (800/1900) to LTE once voLTE is ready). I know they said they would keep CDMA for voice for some time, but in doing so who sits in the more favorable spectrum position.

[johnhere]

I think if AT&T was as aggressive as Verizon with the lte build they would be the carrier to get. The fall back speeds are faster and that's a plus but from a marketing standpoint they need to be more aggressive. I have people ask me all the time if AT&T is planning to build a 4g network.I think both have enough for the immediate future depends on what they do with it.

[zephxiii]

With verizon LTE getting congested it depends on how much they can increase the site density compared to what it is now at 700mhz. They are foreseeing a problem in the future which is why they want to buy AWS.

[ggore]

WTF no they do not need to skip HSPA since if they are installing LTE they are installing equipment that very easily supports HSPA. And given how huge the HSPA capable user base is it would be dumb as hell to skip it. You are better off going HSPA first then LTE later than anything.

And meanwhile, Verizon has had LTE running for years and moved on to LTE-Advanced before AT&T finally gets to that LTE upgrade? Really? I call WTF to that!

[fraydog]

CSFB works either on UMTS or GSM. AT&T certainly could deploy in that manner.

[zephxiii]

And meanwhile, Verizon has had LTE running for years and moved on to LTE-Advanced before AT&T finally gets to that LTE upgrade? Really? I call WTF to that!

It makes zero sense to all out deploy LTE in GSM only areas without HSPA:

A: don't have enough 700/AWS spectrum to cover everywhere
B: Don't have very much 700/AWS spectrum in a number of rural places to begin with
C: Only have AWS in some areas
D: LTE does not support VOICE
E: The % of user that are LTE capable is very small.
F: Nearly the entire subscriber base can immediately make use of HSPA+ whereas they cannot with LTE
G: GSM is being phased out and HSPA is its replacement.
H: LTE hardware is still new and not completely matured
I: You can easily add LTE later on to the equipment installed for HSPA+

Also AT&T has already planned to fully deploy LTE by end of 2013 or whatever and already has plans to move on to LTE Advanced. You are forgetting that VZW needs LTE since EVDO is way over taxed whereas AT&T HSPA is averaging much better on the same spectrum.

In addition HSPA+ works very well in rural areas from what I've seen.

[fraydog]

AT&T will eventually have their hand forced on LTE in rural areas, however. They're going to use HSPA+ when everywhere else in the US carriers have moved to LTE. Even Sprint and T-Mobile will have footprint wide LTE by the end of 2014. So why shouldn't AT&T?

Zeph, you are skating to where the puck is now and not where it's going to be. Actually with Sprint and T-Mobile deploying footprint wide it would be CHEAPER for AT&T to do the same by the end of 2014 and start retiring legacy spectrum rather than continuing HSPA+. AT&T still has massive quantities of PCS they can refarm for one.

[ggore]

It makes zero sense to all out deploy LTE in GSM only areas without HSPA:


Also AT&T has already planned to fully deploy LTE by end of 2013 or whatever and already has plans to move on to LTE Advanced. You are forgetting that VZW needs LTE since EVDO is way over taxed whereas AT&T HSPA is averaging much better on the same spectrum.

In addition HSPA+ works very well in rural areas from what I've seen.

It works well if it's deployed. Which by my guess is to maybe half the US land area. The rest is EDGE, plain-jane 3G or roaming.

"Fully deployed" for AT&T is not the same as for Verizon. AT&T says they can only deploy LTE to 80% of the population and 55% of the land area of the continental US due to spectrum constraints. This is a far cry from the 97% of the population that Verizon will have covered a year from now. I don't know what the land area percentage is, but from their coverage map it looks like they cover about 80% of the land area, which is a far cry from 55%!

Stated conversely, which puts some real numbers to the situation, HSPA will be the best 45% of the land area of the US will ever have from AT&T. And 20% of the population, which equals around 60 million people!

[fraydog]

It works well if it's deployed.

"Fully deployed" for AT&T is not the same as for Verizon. AT&T says they can only deploy LTE to 80% of the population and 55% of the land area of the continental US due to spectrum restraints. This is a far cry from the 97% population for Verizon. I don't know what the land area percentage is, I've never seen that stated, but it is much more than 55% since Verizon is overlaying all their 3G area with LTE and that area is much larger than 55%.

Stated conversely, which puts some real numbers to the situation, HSPA will be the best 45% of the land area of the US will ever have from AT&T. And 20% of the population, which equals around 60 million people!

AT&T has to be more aggressive anyway. The cost of LTE equipment is greatly falling, the cost of maintaining legacy GSM is actually going up, and AT&T can greatly save money as well as provide better service in the long run going straight to LTE.

By the time it takes to hit all the rural areas with HSPA+ VoLTE will be active. Even still GSM fallback is an interim solution that can and should be used.

AT&T also has considerably better spectrum flexibility with LTE that they never possessed with HSPA+. Take that into consideration. They have AWS and PCS in large blocks west of the Mississippi and can make a few key rural acquisitions that could get their 700 MHz footprint into the same ballpark as VZW.

So I think zeph's argument is invalidated completely due to the falling cost of LTE.

[ggore]

The LTE in this area of Oklahoma was placed by Verizon in a 1X roaming area. Verizon has no native system there. The roaming partner only offers 1X service, no 3G. Verizon didn't need to have native 3G there, nor did the partner, evidently, for them to place an LTE setup in that area. So why does AT&T have to have 3G and/or HSPA in an area before they can deploy LTE? Voice calls are not a problem, essentially Verizon is only offering LTE data there, the voice calls will be handled by the roaming partner I assume until VoLTE is rolled out. Evidently it is possible to skip steps in the overall process, and Verizon is perfectly willing to do that, just another way AT&T will be unable to keep up or compete.

Attachment 81797

[fraydog]

It makes zero sense to all out deploy LTE in GSM only areas without HSPA:

A: don't have enough 700/AWS spectrum to cover everywhere
B: Don't have very much 700/AWS spectrum in a number of rural places to begin with
C: Only have AWS in some areas
D: LTE does not support VOICE
E: The % of user that are LTE capable is very small.
F: Nearly the entire subscriber base can immediately make use of HSPA+ whereas they cannot with LTE
G: GSM is being phased out and HSPA is its replacement.
H: LTE hardware is still new and not completely matured
I: You can easily add LTE later on to the equipment installed for HSPA+

Also AT&T has already planned to fully deploy LTE by end of 2013 or whatever and already has plans to move on to LTE Advanced. You are forgetting that VZW needs LTE since EVDO is way over taxed whereas AT&T HSPA is averaging much better on the same spectrum.

In addition HSPA+ works very well in rural areas from what I've seen.

1. AT&T can certify most of their UE to work with 850 and 1900.
2. See 1.
3. See 1.
4. VoLTE deployments are starting full scale next year. It will take that long for AT&T to get HSPA+ everywhere. Why not just move on?
5. In rural areas where devices are LTE only swap their handsets out. Not that many users.
6. LTE handsets will have HSPA+ for a long long time.
7. LTE is the replacement for HSPA+ and is actually cheaper to maintain over a long period. Since it's very nearly the same equipment why not make the initial network investment for a greater long-run return?
8. The new Snapdragon S4 chips support VoLTE AND greatly enhance battery life. The S4 separate baseband (the MSM9615) is good enough on battery life that Apple is putting it in the iPhone. The S4 also brings SR-VCC deployment into play which enables VoLTE calls to fall back to UMTS circuit switched voice.
9. Just get it done then. Also AT&T has more backhaul in their long line system than anyone. AT&T can microwave to a large cluster of cells off of ONE long lines tower and have more than enough backhaul to serve rural areas for a LONG time.

I defended UMTS for a long time. It is time to acknowledge the failures for rural deployments it is and get to a better widget.

[ggore]

AT&T has to be more aggressive anyway. The cost of LTE equipment is greatly falling, the cost of maintaining legacy GSM is actually going up, and AT&T can greatly save money as well as provide better service in the long run going straight to LTE.

By the time it takes to hit all the rural areas with HSPA+ VoLTE will be active. Even still GSM fallback is an interim solution that can and should be used.

AT&T also has considerably better spectrum flexibility with LTE that they never possessed with HSPA+. Take that into consideration. They have AWS and PCS in large blocks west of the Mississippi and can make a few key rural acquisitions that could get their 700 MHz footprint into the same ballpark as VZW.

So I think zeph's argument is invalidated completely due to the falling cost of LTE.

Let's hope so, otherwise it's a complete no-contest.