Sprint TD-LTE 2500/2600mhz Discussion

[scottlloyd3]

_{Has anyone heard anything about the SF bay area going live?  I read it was also one of the first due out.  Also, anyone heard anything about the new hardware required for the ClearWire modems for home internet use?}

[HTCandroid]

Koi the species of fish, are coy "weird" que "what"

 

So I called us weird fish?! hahaha

 

So AJ said "So koi are weird? What?"

[kineticman]

So I called us weird fish?! hahaha

 

So AJ said "So koi are weird? What?"

So you like fish sticks?

[HTCandroid]

So you like fish sticks?

 

Nope! That would make me a gay fish, and I ain't getting gay with no fish! <-----Reference to the "getting gay with kids are here" episode with guest Jennifer Aniston(one of the few episodes I know) haha.

 

Do you like fish sticks?

[WiWavelength]

So I called us weird fish?! hahaha

 

So AJ said "So koi are weird? What?"

 

I see that you figured it out.  You must be a genius...

 

 

 

AJ

[koiulpoi]

So, koi are coy?  ¿Qué?

 

 

AJ

Yes, I am.

[Conan Kudo]

True. I'm just going to say the next year iPhone will support Triband. By then LTE should be (hopefully) deployed nationwide and most of the major markets are above 60% competition

Maybe, but it's unlikely. Normally supporting Bands 7+38 automatically makes it impossible to support Band 41 (they are currently mutually exclusive). And Apple will definitely support those bands before Band 41, because they're being used EVERYWHERE except Japan and the US. Even Canada uses Band 7 right now, with Band 38 being auctioned sometime next year.

[CaryTheLabelGuy]

Maybe, but it's unlikely. Normally supporting Bands 7+38 automatically makes it impossible to support Band 41 (they are currently mutually exclusive). And Apple will definitely support those bands before Band 41, because they're being used EVERYWHERE except Japan and the US. Even Canada uses Band 7 right now, with Band 38 being auctioned sometime next year.

Are you saying that Apple will most likely never support band 41 and 7+38? What if they make a different model sans 7+38 but include B41?

 

 

Sent from my Sprint iPhone 5, not the old one (using Tapatalk 2).

[Conan Kudo]

Are you saying that Apple will most likely never support band 41 and 7+38? What if they make a different model sans 7+38 but include B41?

 

 

Sent from my Sprint iPhone 5, not the old one (using Tapatalk 2).

It wouldn't be worth it for Apple, but it might. I suspect that Apple will content itself with Bands 25 and 26, since LTE TDD might be a bit too hard to implement alongside FDD in the iPhone. Remember, Apple prefers to develop unified devices, and I wouldn't doubt that Apple has been working to develop a single GSM/CDMA2000/UMTS/LTE device to cover most of the world. It's currently possible to support 8 bands on one device, so I could see this happening. As long as LTE TDD is left out, Apple could pull it off.

[tmiw]

Maybe, but it's unlikely. Normally supporting Bands 7+38 automatically makes it impossible to support Band 41 (they are currently mutually exclusive). And Apple will definitely support those bands before Band 41, because they're being used EVERYWHERE except Japan and the US. Even Canada uses Band 7 right now, with Band 38 being auctioned sometime next year.

 

People said for the longest time that Verizon would never get the iPhone because they didn't use GSM, and here we are. But without the chipset support, yeah, there'll probably only be support for two out of the three bands.

[Rawvega]

It wouldn't be worth it for Apple, but it might. I suspect that Apple will content itself with Bands 25 and 26, since LTE TDD might be a bit too hard to implement alongside FDD in the iPhone. Remember, Apple prefers to develop unified devices, and I wouldn't doubt that Apple has been working to develop a single GSM/CDMA2000/UMTS/LTE device to cover most of the world. It's currently possible to support 8 bands on one device, so I could see this happening. As long as LTE TDD is left out, Apple could pull it off.

You certainly could be right. However, I remember reading similar comments over on HoFo that Apple wouldn't support Band Class 10 SMR 800 on the iPhone because Sprint is the only carrier in the world using it. Of course, the iPhone 5 proved that theory false. Trying to predict what Apple is good for amusement, but usually an exercise in futility.

[digiblur]

You certainly could be right. However, I remember reading similar comments over on HoFo that Apple wouldn't support Band Class 10 SMR 800 on the iPhone because Sprint is the only carrier in the world using it. Of course, the iPhone 5 proved that theory false. Trying to predict what Apple is good for amusement, but usually an exercise in futility.

Except one huge difference, adding 800smr support was relatively easy.

[rf40928]

That sounds pretty good in comparison. I thought the Sprint NV LTE sites only had 100 mbps connection for backhaul (with possible upgrade to 1 Gbps)

 

 

 

 

 

 I don't see how.. Only 100 Mb/s ?  If you're downloading at 7 Mb/s on a single phone ( which is typical for LTE ) you'd only have enough bandwidth for a dozen phones on that tower...

[Thomas L.]

Why should it be hard to implement FDD and TDD together? As far as I know the chipsets are designed to allow them to coexist easily and all chipsets support both.

 

Sent from my SPH-L710 using Xparent ICS Blue Tapatalk 2

[Conan Kudo]

Except one huge difference, adding 800smr support was relatively easy.

Correct. SMR and ESMR are adjacent to Cellular band, and use the same duplex gap. Thusly, the Cellular band can be easily extended to cover the frequencies. That's how Band 26 exists.

 

Why should it be hard to implement FDD and TDD together? As far as I know the chipsets are designed to allow them to coexist easily and all chipsets support both.

 

Sent from my SPH-L710 using Xparent ICS Blue Tapatalk 2

 

The problem isn't the baseband chipset. It's the RF frontend components. TDD components are vastly more expensive and require different techniques to properly implement than FDD. Also, smaller phones make it much more difficult to implement both modes. Effectively, a completely separate radio chain is required for TDD. That's also why SV-LTE goes away with Sprint LTE FDD+TDD phones.

[Rawvega]

Except one huge difference, adding 800smr support was relatively easy.

However, if Apple were to add TD-LTE support for China Mobile, why wouldn't it be relatively easy to add TD-LTE support for Sprint?

 

http://seekingalpha.com/article/1138601-china-mobiles-td-lte-technology-expanding-fast-where-to-invest

 

"We plan to adopt multi-mode, multi-band terminals, with basic terminals supporting 5 modes and 10 spectrum bands, and ideally supporting 5 modes and 12 bands. By 2014, we will require that terminals support Band 41," explained Wang Hengjiang, director of the products division of China Mobile Terminal.

[koiulpoi]

 I don't see how.. Only 100 Mb/s ?  If you're downloading at 7 Mb/s on a single phone ( which is typical for LTE ) you'd only have enough bandwidth for a dozen phones on that tower...

A single carrier of 5x5 MHz FD-LTE maxes out at ~37 Mbps. You've usually got this split to (multiplied by) 3 sectors per site, so yeah. 100 Mbps should definitely cut it for now. With the addition of an additional 5x5 LTE Carrier on ESMR 800 MHz, they'll probably need ~200 Mbps to fully serve. Remember, most data is in "bursts"; it doesn't need to be always-on once it's loaded the page for you, so it can serve many more than that with good speeds.

 

What your post illustrates is the inherent shared nature of wireless services. When one person is using something, somebody else isn't, or is slower. This is why things like abuse of unlimited data and unauthorized tethering are so harmful to the network. If one guy is sitting there chewing through all the bandwidth of a single sector/carrier, it harms the experience of others on that site.

[Conan Kudo]

However, if Apple were to add TD-LTE support for China Mobile, why wouldn't it be relatively easy to add TD-LTE support for Sprint?

 

http://seekingalpha.com/article/1138601-china-mobiles-td-lte-technology-expanding-fast-where-to-invest

Apple does not deal with China Mobile currently, and it's not likely it ever will, since Apple would have to incorporate GSM, WCDMA (with CA for HSPA+), TD-SCDMA, CDMA2000, LTE TDD (with MIMO and CA), and LTE FDD (with MIMO). That makes the iPhone impossibly expensive for China Mobile's audience. And unlike China Unicom and China Telecom (who both have customers that actually sign contracts and pay a lot of money), China Mobile's customers are almost entirely prepaid with demands for cheap devices.

[rf40928]

A single carrier of 5x5 MHz FD-LTE maxes out at ~37 Mbps. You've usually got this split to (multiplied by) 3 sectors per site, so yeah. 100 Mbps should definitely cut it for now. With the addition of an additional 5x5 LTE Carrier on ESMR 800 MHz, they'll probably need ~200 Mbps to fully serve. Remember, most data is in "bursts"; it doesn't need to be always-on once it's loaded the page for you, so it can serve many more than that with good speeds.

 

What your post illustrates is the inherent shared nature of wireless services. When one person is using something, somebody else isn't, or is slower. This is why things like abuse of unlimited data and unauthorized tethering are so harmful to the network. If one guy is sitting there chewing through all the bandwidth of a single sector/carrier, it harms the experience of others on that site.

Ok, Im confused. Are we talking 100 MEGABYTES ( MB ) or 100 Megabits ( Mb )... Cuz you're saying 100 Mb/s ( Megabits ) should cut it.. I'm thinking maybe 100 MB/s will cut it.. That's why I questioned the Original Poster..  I get 50 Mb's ( Megabits ) on my home internet download.. You're saying a 5x5 FD-LTE maxes out at -37 Mb/s ( Megabits?? ) or (MB ) Megabytes... ? Huge difference.. That would put your single 5x5 FD-LTE site (servicing a whole area) - using fiber - at less bandwidth - then what I have ( all by myself ) on my single home Road Runner connection ( also using fiber ) that gives me a 50 Mb/s down and 5 Mb/s up all by myself .. There are 8 Megabits ( Mb ) in 1 Megabyte ( MB )... My Road Runner 50 Mb/s download is equal to 6.25 MB/s ... Huge difference..

 

1,000 MB = 1 GB ( the max? for backhaul )...  100 MB is only 1/10th of 1 GB... On the other hand 100 Mb/s is only about 12.5 MB !! And you're saying a 5x5 FD-LTE maxes out at 37 Mb?? No Way

[koiulpoi]

Ok, Im confused. Are we talking 100 MEGABYTES ( MB ) or Megabits ( Mb )... Cuz you're saying 100 Mb/s ( Megabits ) should cut it.. I'm thinking maybe 100 MB/s will cut it.. That's why I questioned the Original Poster..  I get 50 Mb's ( Megabits ) on my home internet download.. You're saying a 5x5 FD-LTE maxes out at -37 Mb/s ( Megabits?? ) or (MB ) Megabytes... ? Huge difference.. That would put your 5x5 FD-LTE servicing a whole area ( using fiber ) at less bandwidth then what I have on my home Road Runner connection that gives me a 50 Mb/s down and 5 Mb/s up all by myself

Megabits. Small B. And yes, that's exactly what I'm saying. Wireless is a shared resource, and it is much more difficult to create more capacity for wireless than it is to run some more coax/fiber. Even the "almighty Verizon" has a 10x10 MHz FD-LTE Network, which maxes out at ~72 Mbps, so 10 concurrent users drop that to 7.2 Mbps each, and 100 drop it to 0.72 Mbps. Again though, you have to multiply the total capacity of the site by 3, as there are typically three sectors per carrier (basically each covering 120 degrees of space around the site). So again, for Sprint, you've got 37 + 37 + 37 = a total theoretical maximum of 111 Mbps, which is why most sites will get by with 100.

 

Which again, I think this very well illustrates that Wireless technology is not the "end all be all" that it is often made out to be, and LTE is not the perfect panacea. There is still nothing that compares to good old wires do your doorstep. To make wireless work, you have to throw a lot of spectrum at it, or you have to split your cells and throw up a ton of sites (or introduce small cells).

[maximus1987/lou99]

Megabits. Small B. And yes, that's exactly what I'm saying. Wireless is a shared resource, and it is much more difficult to create more capacity for wireless than it is to run some more coax/fiber. Even the "almighty Verizon" has a 10x10 MHz FD-LTE Network, which maxes out at ~72 Mbps, so 10 concurrent users drop that to 7.2 Mbps each, and 100 drop it to 0.72 Mbps. Again though, you have to multiply the total capacity of the site by 3, as there are typically three sectors per carrier (basically each covering 120 degrees of space around the site). So again, for Sprint, you've got 37 + 37 + 37 = a total theoretical maximum of 111 Mbps, which is why most sites will get by with 100.

 

Not so fast. Double that for 800 LTE and add some more for the EVDO.

Sprint will quickly need 250mbps at most sites and that's not even counting TDD.

Edited July 26, 2013 by asdf190

[koiulpoi]

Not so fast. Double that for 800 LTE and add some more for the EVDO. Sprint will quickly need 250mbps at most sites and that's not even counting TDD.

I mentioned the addition of an a single 5x5 MHz TD-LTE Carrier on ESMR 800 MHz a couple posts up, but thank you. I neglected to include EV-DO and 1X in my super-vague-near-the-ballpark calculations, as, well, EV-DO is going use up a maximum of 9 Mbps per carrier (assuming 3 sectors), and 1X... I have no idea. Less than 1 Mbps, I'm sure. But yes, 250 Mbps should be good at that point. Let's not forget that Sprint's backhaul contracts are "scalable", meaning they can ask them for more if and when it is needed.

[rf40928]

Not so fast. Double that for 800 LTE and add some more for the EVDO. Sprint will quickly need 250mbps at most sites and that's not even counting TDD.

This makes much more sense..  When you're doing even a speedtest and hitting 16 Mb/s.. I'd ALONE being using much of the bandwidth of a single site.. let alone the 300 people surfing, texting, steaming music, Netflix.. If they eventually  got 250 Mb/s that give you about 31.25 MB/s .. much better

[Steven800]

I think he's talking Megabits. If a cell site had 3 sectors and each one is capable of 37 megabits. You'd need 111 Megabits of backhaul per tower. But since most towers usually don't have people constantly using data,then 100 Megabits should be roughly good enough for each tower. Now I think that there's probably more than that because they are going to need more backhaul for LTE on 800mhz. And way more than that if they decide to put LTE 2600mhz on the same tower as sprint.

 

Edit: I typed this in my phone. Then when hit post saw that you'll beat me to It.

[koiulpoi]

This makes much more sense..  When you're doing even a speedtest and hitting 16 Mb/s.. I'd ALONE being using much of the bandwidth of a single site.. let alone the 300 people surfing, texting, steaming music, Netflix.. If they eventually  got 250 Mb/s that give you about 31.25 MB/s .. much better

I'm not sure if you understand, or maybe I'm not being clear in my explanation... Sprint is using a single 5x5 MHz carrier for their LTE right now. This one carrier has a capacity, and maximum speed, of 37 Mbps. It doesn't matter how much of a pipeline behind it, the 10 MHz of airwaves can't deliver any more than that, period.

 

The reason that you multiply that by three is that Sprint typically splits sites (towers) in to three "sectors". Imagine you've got a circle going out from a cell site, and split it in to three evenly-sized "triangles", so you cover 360 degrees with three 120 degree sectors. Make sense?

 

So, since you have a maximum of 37 Mbps (little b, megabits), times three for three sectors, you get a theoretical maximum capacity of 111 Mbps. You can put 1000 Mbps behind that and it won't make it any faster. A single cell site using 10 MHz (5x5) of FD-LTE literally cannot go any faster.

 

Now, Sprint is also putting another 5x5 on the old Nextel frequency (800 MHz ESMR), so double the capacity right there. That's 20 MHz total, using ~222 Mbps theoretical maximum.