Questions about how to obtain 20X20Mhz.

[Exzir]

Hey,

 

I know you have to have consecutive block access to obtain a 20X20 LTE but could does it have to be like Block A and block D. Here is the info I have ATM.

 

Block A: 1850 - 1865 and 1930 - 1945 (30 MHz) - issued by MTAs

Block B: 1870 - 1885 and 1950 - 1965 (30 MHz) - issued by MTAs

Block C: 1895 - 1910 and 1975 - 1990 (30 MHz, 15 MHz or 10 MHz) - issued by BTAs

Block D: 1865 - 1870 and 1945 - 1950 (10 MHz) - issued by BTAs

Block E: 1885 - 1890 and 1965 - 1970 (10 MHz) - issued by BTAs

Block F: 1890 - 1895 and 1970 - 1975 (10 MHz) - issued by BTAs

Block G: 1910 - 1915 and 1990 - 1995 (10 MHz) - issued by EAs

 

Block A and Block D would be 1845-1870 and 1930- 1950 or would you need consecutive letters like D,E,F and G.

 

I know you need 40 Mhz total spectrum so you would need a 30 and a 10; or 4 10's.

[jefbal99]

You'd need LTE-Advanced to do what you are suggesting. The best Sprint could do in its existing would be a 15x15, but i don't think any of the current phones (expect the iPhone) support more than a 5x5 or 10x10 channel.

 

Sprint will be just fine doing multiple 5x5 channels as they fully utilize their A-F spectrum to compliment the G block and SMR.

[WiWavelength]

I can certainly appreciate your interest in this question if it is academic. I think much the same way. That said, your question is of little practical significance. In PCS 1900 MHz spectrum, LTE 20 MHz FDD (i.e. 20 MHz x 20 MHz) will be exceedingly rare, possibly even nonexistent.

 

First, you have to understand the PCS band plan. For various licensing reasons, it does not follow alphabetical order. See below:

 

PCS A block (30 MHz)

PCS D block (10 MHz)

PCS B block (30 MHz)

PCS E block (10 MHz)

PCS F block (10 MHz)

PCS C block (30 MHz)

PCS G block (10 MHz)

 

To deploy LTE 20 MHz FDD would require 40 MHz contiguous spectrum -- for example, a PCS A+D, D+B, B+E, F+C, or C+G block combo. A 20 MHz disaggregation from a 30 MHz license plus two adjacent 10 MHz licenses would be another possibility, but that is so uncommon as to be almost irrelevant. Even the 30 MHz plus 10 MHz adjacent license combo is relatively rare. Sure, VZW, AT&T, and T-Mobile have some markets in which they hold 40 MHz total PCS spectrum, but only coincidentally is that 40 MHz contiguous. The PCS band was not designed nor auctioned in the 1990s for huge swaths of contiguous spectrum.

 

So, in this decade at least, LTE 20 MHz FDD is not likely to be found outside the AWS 2100+1700 MHz band. Both VZW and T-Mobile have opportunities to deploy 20 MHz x 20 MHz in AWS. A year or two earlier, though, Clearwire will be deploying LTE 20 MHz TDD in its BRS/EBS 2600 MHz spectrum. That is likely to be the first taste of LTE 20 MHz carriers.

 

AJ

[irev210]

Interesting post AJ.

 

Another question:

Three scenarios

5x5 + 5x5 + 5x5 + 5x5 = 4 channels, 1 RRU, 4 line cards needed?

10x10 + 10x10 = 2 channels, 1 RRU, 2 line card needed?

20x20 = 1 channel, 1 RRU, 1 line card needed?

 

Is this the right way of looking at it? Is this why there is a desire to create such wide channels? A lot cheaper (in terms of hardware needed to deploy)?

[S4GRU]

Interesting post AJ.

 

Another question:

Three scenarios

5x5 + 5x5 + 5x5 + 5x5 = 4 channels' date=' 1 RRU, 4 line cards needed?

10x10 + 10x10 = 2 channels, 1 RRU, 2 line card needed?

20x20 = 1 channel, 1 RRU, 1 line card needed?

 

Is this the right way of looking at it? Is this why there is a desire to create such wide channels? A lot cheaper (in terms of hardware needed to deploy)?[/quote']

 

Cards are not prohibitively expensive. But there is some hardware and even a small operational savings. But the savings from having lower frequency spectrum and reduced site count is far better than reducing carriers at existing sites.

 

Robert via Moto Photon Q using Forum Runner

[digiblur]

A nice link explaining the bands, blocks, and channels.

 

http://tk.com/wireless/1900pcs.html

[WiWavelength]

Is this why there is a desire to create such wide channels? A lot cheaper (in terms of hardware needed to deploy)?

 

I doubt it. The desire for greater bandwidth LTE carriers is driven by the lack of disadvantages (pardon the double negative) of greater bandwidth LTE carriers. In other words, if a provider has the contiguous and available spectrum to do 20 MHz FDD, then it probably will do 20 MHz FDD. The one downside that comes to mind is sometimes a handset ERP/EIRP reduction on the uplink due to the greater bandwidth, but even that is somewhat dependent on the number of Resource Blocks assigned to the mobile.

 

AJ

[irev210]

I doubt it. The desire for greater bandwidth LTE carriers is driven by the lack of disadvantages (pardon the double negative) of greater bandwidth LTE carriers. In other words, if a provider has the contiguous and available spectrum to do 20 MHz FDD, then it probably will do 20 MHz FDD. The one downside that comes to mind is sometimes a handset ERP/EIRP reduction on the uplink due to the greater bandwidth, but even that is somewhat dependent on the number of Resource Blocks assigned to the mobile.

 

AJ

 

I just tend to think that by creating smaller carriers, you are creating a more consistent experience for end users (ie, one user can't suck an entire 20x20 channel). If it's not a cost issue, you are only creating more variability in the user experience, which makes no sense.

 

But if you have less hardware, less energy, less capital costs, then the wide channels start to make a lot of sense.

 

Thx for explaining (thx to robert as well)

[iansltx]

I just tend to think that by creating smaller carriers, you are creating a more consistent experience for end users (ie, one user can't suck an entire 20x20 channel). If it's not a cost issue, you are only creating more variability in the user experience, which makes no sense.

 

But if you have less hardware, less energy, less capital costs, then the wide channels start to make a lot of sense.

 

Thx for explaining (thx to robert as well)

 

I dunno about consistency. There's a reason that DOCSIS 3 has channel bonding...more capacity, more consistency. Then again cable companies have speed tiers.

 

The pro to 20x20 LTE is you have tons of capacity in one place, without the need for Carrier Aggregation. The con is that most phones don't support that bandwidth. Heck, most Sprint phones only support 5x5.

[irev210]

I dunno about consistency. There's a reason that DOCSIS 3 has channel bonding...more capacity, more consistency. Then again cable companies have speed tiers.

 

The pro to 20x20 LTE is you have tons of capacity in one place, without the need for Carrier Aggregation. The con is that most phones don't support that bandwidth. Heck, most Sprint phones only support 5x5.

 

I think given that cable companies are stuck with 6MHz channels, they have no choice but to either go to a higher QAM (more dense) or bond channels.

 

Wireless operators seem to have the flexibility to deploy whatever channel width the standard they are deploying supports.

[iansltx]

I think given that cable companies are stuck with 6MHz channels, they have no choice but to either go to a higher QAM (more dense) or bond channels.

 

Wireless operators seem to have the flexibility to deploy whatever channel width the standard they are deploying supports.

 

The implementation is different (4x6 MHz channels or 8x6 MHz channels vs. on a 5x5, 10x10, 15x15 or 20x20 channel) but the reasoning is the same: carriers want a larger pipe for their users if it can be done relatively inexpensively using available resources. D3 channel bonding these days is seamless enough that it might as well be one big 24MHz (or however many DOCSIS channels get bonded, multiplied by 6) downstream channel. Says the person who has had D3 in some form or fashion for maybe three years now.

[WiWavelength]

I think given that cable companies are stuck with 6MHz channels, they have no choice but to either go to a higher QAM (more dense) or bond channels.

 

I have not heard of a move to greater than 256-QAM, even over wired connections. For example, 512-QAM increases modulation complexity 100 percent but increases potential throughput only 12.5 percent. I call that the point of diminishing returns.

 

AJ

[irev210]

I have not heard of a move to greater than 256-QAM, even over wired connections. For example, 512-QAM increases modulation complexity 100 percent but increases potential throughput only 12.5 percent. I call that the point of diminishing returns.

 

AJ

 

 

Thought this was interesting:

 

OFDM will be matched up with low density parity-check (LDPC), a Forward Error Correction (FEC) scheme that takes up less bandwidth than the current Reed-Solomon approach. LDPC will let cable pump out more bits per hertz by utilizing higher orders of QAM modulation, including 1024 QAM and 4096 QAM in both the downstream and the upstream, Schmitt said. (256 QAM is typically used in today's cable downstream.)

To point out how much more efficient the new approach will be, Schmitt noted that it would take 780MHz of spectrum to get 5 Gbit/s using Docsis 3.0, but only 500MHz with Docsis 3.1.

 

http://www.lightreading.com/document.asp?doc_id=226144&site=lr_cable

[bigsnake49]

There are couple of cons to wider channels. If that wider channel card goes out, the whole sector goes out unless you have a standby spare you can automatically provision remotely. if one of your 4 5x5 goes out, you have 3 left. On the handset side, transmission over a wider channel, if allocated the maximum number of subcarriers sucks the battery up.