T-Mobile LTE & Network Discussion

[WiWavelength]

That's a cute thought, to think our FCC could or would at least try to sync bands with the EU, but we're always at war with Eurasia.

 

Nope. Blame the rest of the world that seemingly tries to stick it to the US. Like it or not, the US is still the most important country/market on the planet.

 

AJ

[Fraydog]

 

Nope. Blame the rest of the world that seemingly tries to stick it to the US. Like it or not, the US is still the most important country/market on the planet.

 

AJ

 

On most affairs, I agree wholeheartedly. It would still be nice to see cooperation on getting spectrum in the same block. I don't expect to see it in my lifetime.

[WiWavelength]

On most affairs, I agree wholeheartedly. It would still be nice to see cooperation on getting spectrum in the same block. I don't expect to see it in my lifetime.

 

In this case, we may hold opposite perspectives because I would expect the US to yield on most matters. But not on this matter. Eurasian interests have specifically screwed the US at least twice now.

 

A decade and a half ago, the IMT 2100+1900 MHz band (i.e. W-CDMA/LTE band 1) could not be accommodated in the US because the uplink overlapped with the PCS 1900 MHz band set up here five years earlier. Now, the EU "Digital Dividend" 800 MHz band (i.e. LTE band 20) is not possible in the US because the uplink overlaps with the Cellular 850 MHz band set up here 30 years earlier, not to mention, the downlink would wipe out public safety 800 MHz rebanding, including part of our very own Sprint SMR 800 MHz spectrum.

 

As far as band configuration and airlink technology selection go, the 3GPP and EU act in a protectionist fashion and can kiss my ass.

 

AJ

[HawaiiD]

<br />

[milan03]

Not sure it T-Mobile network is the fastest around, but holy latency!

This is from Houston where they're for some reason 2x5Mhz for now even though they have enough contiguous spectrum for 2x10Mhz.

[ericdabbs]

Not sure it T-Mobile network is the fastest around, but holy latency!

This is from Houston where they're for some reason 2x5Mhz for now even though they have enough contiguous spectrum for 2x10Mhz.

 

Its easy to think you are the fastest when the Tmobile LTE network is not loaded. Those speeds are impressive but Sprint had similar LTE speeds when the first 4 markets were released.

[milan03]

Its easy to think you are the fastest when the Tmobile LTE network is not loaded. Those speeds are impressive but Sprint had similar LTE speeds when the first 4 markets were released.

As I said, I've seen better speeds out there, T-Mobile's are certainly not that impressive, but I've never seen a better latency on any US carrier. I think that latency is remarkable, most likely due to improved core and antenna integrated radios.

[S4GRU]

 

Its easy to think you are the fastest when the Tmobile LTE network is not loaded. Those speeds are impressive but Sprint had similar LTE speeds when the first 4 markets were released.

 

That is the best latency I have ever seen on LTE before. The best I've ever had on Sprint LTE is high 30's, and that's uncommon. Most of my results were 50-60ms. And I've never had a VZW LTE 750 ping less than 60ms. In fact, I commonly get around 100ms on VZW. And can be over 200ms on a hotspot.

 

Robert via Nexus 7 with Tapatalk HD

[WiWavelength]

Well, when T-Mobile inevitably starts advertising that it has the "fastest" LTE network, that may truly reflect speed -- as in distance ÷ time.

 

 

AJ

[irev210]

That is the best latency I have ever seen on LTE before. The best I've ever had on Sprint LTE is high 30's, and that's uncommon. Most of my results were 50-60ms. And I've never had a VZW LTE 750 ping less than 60ms. In fact, I commonly get around 100ms on VZW. And can be over 200ms on a hotspot.

 

Robert via Nexus 7 with Tapatalk HD

 

I got into the 20's and low 30's when LTE first launched in Boston.

 

Now, latency is way up and upload is faster than download It's pretty bad how quickly the sites are loading up.

 

To be fair, most sites in Boston aren't on yet, leaving the few that are to take a lot more load, but Sprint did "officially" launch the market...

 

I have to admit, I am pretty excited to see what T-Mobile can do with LTE in Boston. I view the DC-HSPA+ network as far superior to Sprint's current LTE network (in Boston).

[S4GRU]

 

I got into the 20's and low 30's when LTE first launched in Boston.

 

Now, latency is way up and upload is faster than download It's pretty bad how quickly the sites are loading up.

 

They need the density. That's for sure. Especially in a place like Central Boston.

 

Robert via Nexus 7 with Tapatalk HD

[WiWavelength]

I view the DC-HSPA+ network as far superior to Sprint's current LTE network (in Boston).

 

DC-HSPA+ also requires twice the spectrum bandwidth of 5 MHz FDD LTE. T-Mobile is using spectrum profligately, just throwing everything and the kitchen sink out there and seeing what sticks.

 

AJ

[irev210]

DC-HSPA+ also requires twice the spectrum bandwidth of 5 MHz FDD LTE. T-Mobile is using spectrum profligately, just throwing everything and the kitchen sink out there and seeing what sticks.

 

AJ

 

Yeah, but they got the spectrum to burn. They started deploying DC-HSPA+ back in 2011 before LTE was really a possibility for them. Might as well do something with it, heh. I am sure that they'll eventually refarm to LTE. Until then, why not evolve WCDMA?

 

It's pretty well know that T-Mobile has more spectrum per customer than any other carrier - why not use and refarm later?

[ericdabbs]

I wonder if latency is better on average between FDD-LTE or TD-LTE.

[WiWavelength]

Yeah, but they got the spectrum to burn.

 

The pipsqueak of the major carriers should not have "spectrum to burn." That is poor management of a public resource.

 

With consolidation in the industry at current levels, competition for spectrum resources is no longer appropriate. Each carrier needs to be allocated spectrum -- each according to its need.

 

AJ

[Deval]

The pipsqueak of the major carriers should not have "spectrum to burn." That is poor management of a public resource.

 

With consolidation in the industry at current levels, competition for spectrum resources is no longer appropriate. Each carrier needs to be allocated spectrum -- each according to its need.

 

AJ

 

Which makes me feel so disappointed that if the Metro/T-Mobile deal goes through, there will be no required spectrum divestiture.

[milan03]

DC-HSPA+ also requires twice the spectrum bandwidth of 5 MHz FDD LTE. T-Mobile is using spectrum profligately, just throwing everything and the kitchen sink out there and seeing what sticks.

 

AJ

Absolutely agreed. And especially in markets with 30Mhz of AWS and HSPA+21 already deployed in PCS, having 20Mhz dedicated to HSPA+42 and only 10Mhz to LTE, makes little sense to me. You can have about the same capacity with two HSPA+21 channels in two different bands.

Not to mention that the download throughput in a 5Mhz FDD layer is on par with HSPA+42.

[Fraydog]

I suspect that in modernized markets, that T-Mobile will eventually get AWS HSPA down to a single channel. It's better to run DC-HSPA in PCS IMO. Running DC-HSPA in both PCS and AWS

[irev210]

The pipsqueak of the major carriers should not have "spectrum to burn." That is poor management of a public resource.

 

With consolidation in the industry at current levels, competition for spectrum resources is no longer appropriate. Each carrier needs to be allocated spectrum -- each according to its need.

 

AJ

 

But how do you define need? If you prevent carriers from taking risk, they have no opportunity to offer innovative wireless products (like DC-HSPA+). Looking back, you could say that it was/is an inefficient use of spectrum but what about an inefficient use of capital? That's the bigger discussion. LTE was likely cost prohibitive relative to DC-HSPA+ when T-Mobile decided to move towards evolving HSPA+ back in 2011.

 

Look at clearwire - they hardly use any of their spectrum. Instead of letting it fallow, why not put a less efficient technology on it today at a price point that makes sense? Offer a service that is a competitive game changer? Isn't that the whole reason Softbank had interest in the US market? Offer some disruptive service options that big red/big blue can't match? They don't need the spectrum today but to steal marketshare from verizon and AT&T they will need the spectrum tomorrow.

 

The idea of perfectly allocating spectrum to carriers is idealistic. I love the idea, but in a practical sense, it is just impossible. Are you going to say each carrier has to offer the same 5GB data plan and same 5Mbit data speed to keep everything on parity? Needs change. What if T-Mobile grows and needs more spectrum? What if all of them NEED to deploy faster broadband? What sort of return on investment should each carrier be allowed to have? Every day subscribers jump from one carrier to another. Would we have an annual FCC Spectrum Re-balance?

 

The only situation that's problematic is when carriers just buy spectrum so others can't use it.

 

At the end of the day "need" is impossible to define. Trying to say how much spectrum someone needs is like saying how much fried chicken someone should be allowed to eat. Competition requires risk, risk requires someone to take a chance on buying more spectrum than they need to offer a service at a price point a competitor can't.

[WiWavelength]

But how do you define need?

 

It is not as difficult as you make it out to be. See two of my previous posts (in this thread):

 

http://s4gru.com/ind...post__p__117648

http://s4gru.com/ind...post__p__117682

 

AJ

[irev210]

It is not as difficult as you make it out to be. See two of my previous posts (in this thread):

 

http://s4gru.com/ind...post__p__117648

http://s4gru.com/ind...post__p__117682

 

AJ

 

 

Allow each carrier to buy spectrum proportionally according to its market share. Then, the carriers can compete on factors that really matter: coverage, service, policies, and price.

 

Coverage, service, policies, and price are all heavily related to what spectrum a carrier owns. Are you going to take 22MHz of C block and cut it up 4 ways? What if a new carrier wants to join the business - they have zero customers... should they not be granted any spectrum since they have no market share?

 

Spectrum would be leased from the public to wireless carriers on a yearly basis, then reapportioned annually. Unlike auctions, which create periodic windfalls, this would ensure a steady stream of revenue to the Treasury. And my understanding is that Japan manages its wireless spectrum this way.

 

Leases sound like a good idea until ultimately lease costs get passed down to customers and it ends up being a tax. They already tax your wireless bill; If you want to ensure a steady stream of income to the Treasury, I suppose you can increase the tax but that would require congress to act. Good luck with that.

 

As far as reallocating spectrum annually, you've said yourself many times about the limitations on how many bands can be built into a phone. Also, what about pricing for service? I think before a carrier makes billions in capital investments, they want to make sure the infrastructure will be there to realize those gains. If Sprint deployed antennas, base stations, radios for ESMR and then the FCC said "well, sorry, you aren't using it very well" perhaps Sprint wouldn't take the risk with Network Vision.

 

A prime example is light squared. They took a LOT of risk, bought a LOT of spectrum that they couldn't end up doing squat with and blew through billions doing it. It was a big risk, could have paid off, ended up not. It was LightSquared that decided to take the risk. You could say the same thing with Dish - they took the risk with the S-Band, ended up working out well for them.

[Deval]

 

 

A prime example is light squared. They took a LOT of risk, bought a LOT of spectrum that they couldn't end up doing squat with and blew through billions doing it. It was a big risk, could have paid off, ended up not. It was LightSquared that decided to take the risk. You could say the same thing with Dish - they took the risk with the S-Band, ended up working out well for them.

 

Not sure how well S-band has worked out for Dish though?

[ericdabbs]

The investment in the S-band has worked out for Dish so far in that it was able to get FCC approval for LTE use and got 3GPP approval which is a win in itself. This makes the S-band spectrum valuable. Whether Dish decides to use it build out a LTE network or sell the S-band spectrum, Dish will have made a profit on it.

[Mobilesolutions]

The pipsqueak of the major carriers should not have "spectrum to burn." That is poor management of a public resource.

 

With consolidation in the industry at current levels, competition for spectrum resources is no longer appropriate. Each carrier needs to be allocated spectrum -- each according to its need.

 

AJ

 

T-mobile may be the smallest now but with some of their recent Ventures, Solavei & "uncarrier" stance i expect to see some growth for T-mobile & perhaps an interesting new twist in the wireless industry; considering they do have deploy-able spectrum to grow with.

[Fraydog]

Didn't Legere say something to the effect of "our Internet isn't congested come use it?" I don't have the exact quote.