PCS H Block Spectrum discussion (was "Draft Rules for H Block Auction Set by FCC"

[Conan Kudo]

Once sprint deploys the H block does that mean their existing 5x5 on 1900 goes to 10x10 on 1900 and will that double the data speeds?

With the current gear, that is not possible. The only way for Sprint to go to 10 MHz FDD with G+H is to upgrade the radio hardware (or get brand new radio hardware) to support the "multiple frequency band indicator" (otherwise known as MFBI) feature that allows Sprint to declare 5 MHz FDD for Band 25 and 10 MHz FDD for a newly defined extended PCS A-H band.

 

Existing devices with Band 25 will only see 5 MHz FDD until the end of their days. Devices with the new band will support 10 MHz FDD with G+H.

[Nickel]

With the current gear, that is not possible. The only way for Sprint to go to 10 MHz FDD with G+H is to upgrade the radio hardware (or get brand new radio hardware) to support the "multiple frequency band indicator" (otherwise known as MFBI) feature that allows Sprint to declare 5 MHz FDD for Band 25 and 10 MHz FDD for a newly defined extended PCS A-H band.

 

Existing devices with Band 25 will only see 5 MHz FDD until the end of their days. Devices with the new band will support 10 MHz FDD with G+H.

What is the 3GPP going to do after band 31 and 32? They're be out of FDD bands unless they start again at 50 or something.

[Conan Kudo]

What is the 3GPP going to do after band 31 and 32? They're be out of FDD bands unless they start again at 50 or something.

There are a few reserved band numbers in place already, such as 15 and 16. However, devices don't strictly use band numbers. They rely on {U,E}ARFCNs, which band numbers are merely a construct of {U,E}ARFCN groups for frequencies. Renumbering is certainly possible, and increasingly likely, too.

[Nickel]

There are a few reserved band numbers in place already, such as 15 and 16. However, devices don't strictly use band numbers. They rely on {U,E}ARFCNs, which band numbers are merely a construct of {U,E}ARFCN groups for frequencies. Renumbering is certainly possible, and increasingly likely, too.

Renumbering seems like it would be a huge mess, not for the people trying to figure out what bands their phones support.

[bigsnake49]

Supplemental downlink for just about anything -- PCS, AWS, Cellular, etc.  But the PCS/AWS-2 H block holder would have the most interesting opportunity, since it could potentially lease or license, say, the lower 10 MHz of the AWS-4 2000-2020 MHz segment for an asymmetric carrier aggregated 5 MHz x 15 MHz LTE carrier.

 

AJ

 

Or take all 25 (20MHz of 2000-2020MHz +5MHz of PCS H) for an assymetric contiguous aggregated band.

[Conan Kudo]

Renumbering seems like it would be a huge mess, not for the people trying to figure out what bands their phones support.

The renumbering would likely only affect LTE TDD bands, as they would probably get pushed down to the 60s.

[Nickel]

The renumbering would likely only affect LTE TDD bands, as they would probably get pushed down to the 60s.

Ah that makes sense.

[bigsnake49]

Or take all 25 (20MHz of 2000-2020MHz +5MHz of PCS H) for an assymetric contiguous aggregated band.

Or 30 MHz if you count PCS G.

[gnoj]

Dish is interested in getting its AWS-4 2000-2020 MHz uplink converted to downlink.  If so, then the conflicts with the PCS/AWS-2 H block downlink go away.  That could make the H block more valuable, as it would be part of a 25 MHz swath of contiguous spectrum that could be used for supplemental downlink.

 

AJ

 

According to Tim Farrar's post, the FCC's decision on converting that segment of AWS-4 from uplink to downlink will occur in December, before the PCS H auction. Either way, they should know by January whether or not they want to bid.

 

Yeah, that's true. I bet since Dish has relented a little bit the H block gets tacked on to PCS, which means yet another new band. Sigh.

 

I wish that the FCC would just mandate that handset makers always use the newer superset bands in place of the more limited ones (i.e. EPCS Band 25 instead of PCS Band 2, 700 MHz Band 12 instead of Band 17), so that whatever band number they assign for PCS A-H is all anybody has to worry about going forward.

 

Once sprint deploys the H block does that mean their existing 5x5 on 1900 goes to 10x10 on 1900 and will that double the data speeds?

 

It would be possible, but the high cost of doing so makes it highly unlikely. All existing LTE handsets (and apparently some upcoming ones as well...) only use Band 25, which is PCS A-G. Furthermore, many (mostly Samsung's handsets), are only approved for 5 MHz wide FDD rather than also supporting 10 MHz channels. So assuming Sprint doesn't want to pay to replace everybody's LTE handsets with newer ones a la Wi-Max, in order to combine G+H into a 10x10 they would have to: 1.) Make sure Samsung and other phone OEM's include support for 10 MHz channels in addition to switching over from Band 25 to the new EPCS A-H band, 2.) Refarm a 5x5 slice of existing PCS A-F spectrum from 1x/Ev-DO to LTE, which may not be possible to do in every market, and 3.) Replace whatever site equipment can't be retrofitted to support PCS H (I've heard the new NV RRUs may support it, but not the antennas).

 

If none of the other spectrum covered by Band 25 is refarmed, then, as Neal mentioned, Sprint would need to modify their site equipment to support MFBI.  Incidentally, that is something AT&T is currently having to do, since they were stubborn and refused to use Band 12 for their LTE from the start.

 

Since carrier aggregation's high power draw is likely to eventually be improved, it probably makes more sense to deploy another 5x5 carrier on PCS H, and when the time comes aggregate G and H together to form a pseudo-10x10 channel, which would have nearly the same performance as a true 10x10 channel. That way all current handsets will continue to happily use the 5x5 G they're using now, and new handsets will have access to G+H in addition to 800 and 2600.

 

Since all new phones that support PCS H will also presumably be tri-band, whichever path Sprint decides to take (two CA 5x5's vs a true 10x10) wouldn't do anything to help the speeds for existing LTE devices beyond the benefit of off-loading that will already be helped along by 800/2600 LTE. Average (but not peak) speeds for current devices will only improve if Sprint adds a second 5x5 carrier in PCS A-F in addition to the existing G block carrier.

[bigsnake49]

Since carrier aggregation's high power draw is likely to eventually be improved...

 

The power draw of a 10Mhz channel (assuming of cource nobody else is using the channel) will always be a problem. So will a 20MHz channel or a 40Mhz channel. That's why on the handset side, it makes no sense to have carrier aggregation.

[gnoj]

I wonder what would happen if Sprint decides it's not worth it to spend over $1.5B on this spectrum (sinking that money into NV instead), and Dish doesn't change it's mind about not participating. Smaller carriers probably wouldn't have that much to spend either. So if no one's willing to fork over that much for the spectrum, they would have to hold another auction with a lower reserve price, right? The FCC would have to give in at a lower price some point since they're mandated by Congress to license it by February 2015.

 

The power draw of a 10Mhz channel (assuming of cource nobody else is using the channel) will always be a problem. So will a 20MHz channel or a 40Mhz channel. That's why on the handset side, it makes no sense to have carrier aggregation.

 

I take it you mean 2 aggregated 5x5's will draw more power than a single 10x10? Because all the other carriers are getting by just fine with channels wider than 5 MHz, as will Sprint soon with 20 MHz TDD.

[bigsnake49]

I wonder what would happen if Sprint decides it's not worth it to spend over $1.5B on this spectrum (sinking that money into NV instead), and Dish doesn't change it's mind about not participating. Smaller carriers probably wouldn't have that much to spend either. So if no one's willing to fork over that much for the spectrum, they would have to hold another auction with a lower reserve price, right? The FCC would have to give in at a lower price some point since they're mandated by Congress to license it by February 2015.

 

 

I take it you mean 2 aggregated 5x5's will draw more power than a single 10x10? Because all the other carriers are getting by just fine with channels wider than 5 MHz, as will Sprint soon with 20 MHz TDD.

 

Try downloading a really big file on a 10MHz empty channel and get back to me with your battery draw.