Non-G-Block PCS LTE Carriers

[Yuhfhrh]

What was the Verizon LTE handset penetration rate 2 years in to their LTE deployment?

 

 

Sent from Josh's iPhone 5S using Tapatalk 2

 

I don't know, but they only had 5% after 1 year.

 

http://s4gru.com/index.php?/topic/282-verizon-wireless-only-5-of-users-on-lte-1-year-after-lte-debut/

[WiWavelength]

I don't know, but they only had 5% after 1 year.

 

Only 5 percent penetration after a year?  I think that is called just a "tip dip."

 

 

AJ

[iansltx]

For what it's worth, Sprint was selling phones with 1x in SMR before they were selling phones with LTE. So my guess is that the majority of Sprint phones at this point can use 1x800 service if it's available. That in and of itself may free up a 1x carrier worth of capacity where more than one is being used in 1900.

 

Now, counting guard bands, three contiguous 1xRTT or 1xEV channels take up 5x5 MHz. You can stack another four in there to bring the total to 10x10. Sprint has various specturm configurations, but 5x5 + 10x10 and 15x15 are relatively common. It's in Sprint's interest to get down to three 1x + EvDO carriers in PCS, then make sure they're contiguous with each other, so that in either case they've got room to drop a couple 5x5 LTE carriers in band 2 territory. That's partially why you're seeing pretty cheap off-contract LTE devices on Boost Mobile and Virgin Mobile...you don't want phones using EvDO when they could be moving the network toward lower EvDO traffic.

[WiWavelength]

Now, counting guard bands, three contiguous 1xRTT or 1xEV channels take up 5x5 MHz. You can stack another four in there to bring the total to 10x10. Sprint has various specturm configurations, but 5x5 + 10x10 and 15x15 are relatively common. It's in Sprint's interest to get down to three 1x + EvDO carriers in PCS, then make sure they're contiguous with each other, so that in either case they've got room to drop a couple 5x5 LTE carriers in band 2 territory.

 

I was doing some brainstorming on this the other day.  Even in markets where Sprint holds only 20 MHz of PCS A-F block spectrum, it may still be able to refarm soon for a second LTE carrier -- if that 20 MHz is contiguous and if Sprint is willing to rejigger the guard bands and carrier channel assignments to retain a fourth CDMA2000 carrier.

 

Now, going forward, I will speak in FDD terms.  So, that contiguous 20 MHz is 10 MHz FDD.  And Ian's scenario of three CDMA2000 carriers would look like this:

 

[ 0.625 MHz guard | 1.25 MHz CDMA2000 | 1.25 MHz CDMA2000 | 1.25 MHz CDMA2000 | 0.625 MHz guard | 5 MHz LTE ]

 

That would be the conventional configuration.  The problem is which gets the short shrift -- CDMA1X or EV-DO?  One or the other would be left with only a single carrier.

 

But let us say that Sprint wanted to get more aggressive.  No rule nor technical standard requires 0.625 MHz guard bands for CDMA2000.  That is just a product of CDMA2000 carrier bandwidth and PCS block sizes.  The math works out only two ways:  zero guard bands or 0.625 MHz guard bands.

 

Also, remember that a 5 MHz FDD LTE carrier is actually 4.5 MHz FDD -- it already has 0.25 MHz guard bands built in.  Add up two 0.625 MHz external guard bands with two 0.25 MHz internal guard bands, and Sprint has up to an additional 1.75 MHz.

 

So, take that additional 1.75 MHz and put it in play:

 

[ 0.2 MHz guard | 1.25 MHz CDMA2000 | 1.25 MHz CDMA2000 | 1.25 MHz CDMA2000 | 1.25 MHz CDMA2000 | 0.2 MHz guard | 4.5 MHz LTE | 0.1 MHz guard ]

 

There it is.  By rearranging carriers and guard bands plus updating PRLs, Sprint could keep four CDMA2000 carriers and deploy a secondary 5 MHz FDD LTE carrier -- all in a PCS A-F block 20 MHz contiguous market, such as Atlanta, Miami, San Francisco, etc.

 

AJ

[cdk]

How many markets are there where Sprint has only 20 MHz of PCS (and are any with less now)? How many of those markets are non-contiguous?

[WiWavelength]

How many markets are there where Sprint has only 20 MHz of PCS (and are any with less now)? How many of those markets are non-contiguous?

 

Off the top of my head, some major markets where Sprint holds 20 MHz of PCS A-F block spectrum that is not contiguous nor adjacent are Cincinnati, Houston, and Memphis.  This plan would not be deployable in those markets.

 

AJ

[JWMaloney]

How many markets are there where Sprint has only 20 MHz of PCS (and are any with less now)? How many of those markets are non-contiguous?

I can think of at least one nearby.

 

AJ

That band plan looks a lot like the one you and I were discussing recently for the 10x10 of the A block Sprint owns in Baton Rouge. If you recall, it looked like this:

 

[ 1.2 guard | 1.25 CDMA 1x | 1.25 EV-DO | 1.25 EV-DO | 5 LTE | 0.05 unused ]

 

I structured it that way based on the statement here that operation "near the edges of the service is forbidden in 1.2 MHz guard bands." Is that no longer the case?

[WiWavelength]

I structured it that way based on the statement here that operation "near the edges of the service is forbidden in 1.2 MHz guard bands." Is that no longer the case?

 

No, the content on that CDG page is vague and probably at least 15 years old.  The "1.2 MHz guard bands" cannot possibly be FDD.  Otherwise, PCS 0025 would be a prohibited channel assignment, yet Sprint uses it for EV-DO on some sites in the Kansas market.  Additionally, W-CDMA in the PCS A block could not be deployed at UARFCN 412.  But that is not true, either.

 

So, the "1.2 MHz guard bands" might refer to total, rather than FDD, just rounding off the 1.25 MHz sum of the 0.625 MHz guard bands in both uplink and downlink to "1.2 MHz guard bands."  However, as I have noted, 0.625 MHz guard bands are not mandated.  Instead, that number is the only logical possibility for CDMA2000 deployed in multiples of 5 MHz FDD blocks, making 0.625 MHz guard bands a de facto standard but not a de jure standard.

 

AJ

[dedub]

can these types of refarming be done remotely or does require new equipment or on-site re-adjustment of existing equipment?

[JWMaloney]

can these types of refarming be done remotely or does require new equipment or on-site re-adjustment of existing equipment?

 

It depends largely upon the hardware configuration at the site. At sites with sufficient hardware, they should be able to do it remotely.

[xdfgf]

I think I just saw in signalcheck on my phone LTE on 1922.5 for the up channel. I usually have b26 or 1912.5. Went to get a screenshot and it switched back to 800. I never thought I would get mad because I was on b26.

[WiWavelength]

I think I just saw in signalcheck on my phone LTE on 1922.5 for the up channel. I usually have b26 or 1912.5. Went to get a screenshot and it switched back to 800. I never thought I would get mad because I was on b26.

 

There is no 1922.5 MHz in the PCS A-G uplink or downlink.  Those ranges are 1850-1915 MHz for uplink, 1930-1995 for downlink. So, you just saw a SignalCheck bug.  Have you updated the app recently?  Either way, report it to Mike.

 

AJ

[cdk]

Stupid question time, are there any markets where Sprint could deploy a 10 MHz PCS carrier (as a second PCS carrier)?

[newyork4me]

Stupid question time, are there any markets where Sprint could deploy a 10 MHz PCS carrier (as a second PCS carrier)?

 

Yes, Chicago and St. Louis. They should have done so too...the wider carrier is more efficient, and the customers with the (crappy) Samsung phones that only support 5MHz channels are in the minority (and will only get smaller).

[anthony.spina97]

Yes, Chicago and St. Louis. They should have done so too...the wider carrier is more efficient, and the customers with the (crappy) Samsung phones that only support 5MHz channels are in the minority (and will only get smaller).

Sure, they may be in the minority, but if Sprint were to make their current 5x5 carrier in Chicago into a 10x10 carrier, all of those "crappy" Samsung devices would instantly go from having LTE to only being able to connect to 3G. Sprint is already hated enough by most people in the country; how much more do you think they will be hated if they all of a sudden make it so that all those Samsung devices couldn't connect to LTE? The media would lose their minds if that were to happen.

 

-Anthony

 

Edit: Removed St. Louis because AJ stated a 10x10 wouldn't be possible there.

[WiWavelength]

Yes, Chicago and St. Louis. They should have done so too...the wider carrier is more efficient, and the customers with the (crappy) Samsung phones that only support 5MHz channels are in the minority (and will only get smaller).

 

No, not in St. Louis.  The PCS A-F block spectrum gained from USCC is 10 MHz (5 MHz FDD), and it is not contiguous with existing Sprint spectrum.  So, please do not make that assertion.

 

Additionally, for the record, Sprint variant Samsung devices have been receiving Class II filings at the FCC OET for other LTE carrier bandwidths.  Those devices, though, will likely require software updates that may not trickle down to all users.

 

And the claim that wider LTE carrier bandwidths are more efficient is oft spoken but unproven.  Back it up with evidence.  Otherwise, it is mainly there to stimulate the peak speed e-penis.

 

AJ

[travismheim]

 

 

And the claim that wider LTE carrier bandwidths are more efficient is oft spoken but unproven. Back it up with evidence. Otherwise, it is mainly there to stimulate the peak speed e-penis.

 

AJ

I've read on here a few times that wider LTE carriers provide better performance at the outer edges of coverage. Is that not the case, and the better question to ask is what does that even mean? I thought it meant the wider LTE carrier will provide more usability at lower signal levels, ie further away from the tower or with more obstructions.

 

An example would be the VZW LG G2 I was issued for work. LTE on that device seems to work at much lower signal levels, -130 to -135. Where as when my Nexus 5 on Sprint reaches around -125 and LTE becomes mostly unusable and it switches back to evdo. Is that a characteristic of the wider 10mhz FDD carrier VZW has deployed vs 5mhz, or the frequency 700 vs 1900, or something else all together?

[S4GRU]

I don't get usable VZW service beyond -120/-125 RSRP on AWS or 750MHz in my area, or down in New Mexico. 750MHz runs 10MHz channels and AWS runs 15MHz channels around here.

 

In my experiences, the wider channel doesn't make a stronger signal at the cell edge. It just performs better.

 

Robert via Samsung Note 8.0 using Tapatalk Pro

[travismheim]

At home in my basement where my office area is, the G2 will hang on to LTE around -130 and I get 2-3 Mbps down, 1-2 up with ping in the 100-150ms range. If it gets much lower it will drop to evdo. Band 4 is available here, but I don't know what the channel size is. I've only seen two reports of it from LTE discovery in the log. I didn't mean the wider channel makes for a stronger signal, just that the signal will be usable and the phone holds on to it at lower signal levels. I don't even know if that's true, it just seems that way from my experience. It could be some other factor which is what I was asking about.

[travismheim]

Another example, since I'm probably not being very clear would be with Sprint's band 41. I've seen decent speed test results at -138 rsrp. Is that because of the larger 20mhz channel size, or something else? What I'm inferring is the larger channel size would offer slightly better coverage over a smaller width carrier since it works at lower signal levels. Is that correct?

[SpenceSouth]

At home in my basement where my office area is, the G2 will hang on to LTE around -130 and I get 2-3 Mbps down, 1-2 up with ping in the 100-150ms range. If it gets much lower it will drop to evdo. Band 4 is available here, but I don't know what the channel size is. I've only seen two reports of it from LTE discovery in the log. I didn't mean the wider channel makes for a stronger signal, just that the signal will be usable and the phone holds on to it at lower signal levels. I don't even know if that's true, it just seems that way from my experience. It could be some other factor which is what I was asking about.

Verizon's band 4 rollout seems to be very deliberate and targeted in their launched cities. If I go into the heavily trafficked areas of Orlando I can usually find band 4 pretty regularly on my girlfriends device, but outside of those areas it seems very hit or miss. Makes me appreciate Sprint's rollout more and more.

 

I shouldn't assume you are using Verizon as well but it seems to fit what I've observed.

 

Sent from my Nexus 5 using Tapatalk

[S4GRU]

Another example, since I'm probably not being very clear would be with Sprint's band 41. I've seen decent speed test results at -138 rsrp. Is that because of the larger 20mhz channel size, or something else? What I'm inferring is the larger channel size would offer slightly better coverage over a smaller width carrier since it works at lower signal levels. Is that correct?

The B41 experience is highly dependent on device and signal quality. Some devices seem to hold down to -140 RSRP where some will not hardly go beyond -125 in the same area.

 

Also, the quality of the signal is hugely important. Not all -140 RSRP signals are equal. Two exactly the same devices at -140 RSRP three miles out from a tower, one inside the house may connect, but one in the backyard away from the tower may not. Because the one indoors may be shielded from more interference that keeps the signal quality higher and the one outdoors may have more interference. Enough to keep it from connecting with the weak signal.

 

Also, a LOS -140 RSRP is typically higher quality than a -140 that goes through walls or dense foliage. Signal strength is just one factor.

 

There are just so many variables, it is all but impossible to draw distinct arguments about signal strength improvements with wider channels based solely on B41 observation. But one thing that is fully replicable is that wider channels have higher throughput at the edge of service than narrower channels, when all things are equal. But increased MIMO, like 4x2, or radios like 8T8R, improve edge of cell performance dramatically too.

 

Robert via Samsung Note 8.0 using Tapatalk Pro

 

EDIT: another thought about B41, is because of the weaker propagation, signals are much less likely to overlap, especially with the current WiMax overlay. Less overlapping means less interference at the cell edge. This also is a condition that would help cling to really weak signals. Because there will be less interference reducing signal quality. As more and more B41 goes live, the cell edge interference will increase though.

[travismheim]

Verizon's band 4 rollout seems to be very deliberate and targeted in their launched cities. If I go into the heavily trafficked areas of Orlando I can usually find band 4 pretty regularly on my girlfriends device, but outside of those areas it seems very hit or miss. Makes me appreciate Sprint's rollout more and more.

 

I shouldn't assume you are using Verizon as well but it seems to fit what I've observed.

 

Sent from my Nexus 5 using Tapatalk

So far I've only seen Verizon's band 4 near the state fairgrounds. I don't know exactly where it's coming from, since I only catch it when driving past on the highway to and from work. It makes sense to deploy it there for events and such, but I also like the way Sprints deployment is laid out for more uniform coverage. Wimax had most of Syracuse and surrounding areas covered, I expect band 41 will be very similar which will be nice.

 

I don't want to quote a large post, but thank you for the explanation about band 41, and the lower signal levels. It makes more sense now. I can usually grasp most things, but the inner workings of RF can be confusing for me at times.

[newyork4me]

Sure, they may be in the minority, but if Sprint were to make their current 5x5 carrier in Chicago into a 10x10 carrier, all of those "crappy" Samsung devices would instantly go from having LTE to only being able to connect to 3G. Sprint is already hated enough by most people in the country; how much more do you think they will be hated if they all of a sudden make it so that all those Samsung devices couldn't connect to LTE? The media would lose their minds if that were to happen.

 

-Anthony

 

Edit: Removed St. Louis because AJ stated a 10x10 wouldn't be possible there.

This isn't true. The Samsung devices would still connect to the PCS G-block spectrum. They just wouldn't connect to the additional carriers.

[anthony.spina97]

This isn't true. The Samsung devices would still connect to the PCS G-block spectrum. They just wouldn't connect to the additional carriers.

No, they wouldn't. If the current 5x5 carrier was widened to a 10x10 carrier, the devices in question wouldn't connect at all. The only way to add the capacity without leaving said devices in the dark is to add a second 5x5 carrier. Which means there would be 2 5x5 carriers rather than 1 10x10 carrier.

 

-Anthony