reedacus25

 

On 5/14/2021 at 6:24 PM, mdob07 said:

So I wonder if this Sprint keep site conversion is the same setup just with different antenna. There's only 1 Nokia radio, so its either 600/700 only or 1900/2100 only + 2500. It wouldn't make sense to be 600/700 with 2500 without SA N41 because it would only be able to do B41 LTE without 1900/2100 to serve as anchor for NSA N41. The Sprint B25 radio was retained and is connected to the antenna and is currently the only technology broadcasting. This is also a very dense area for T-Mobile. 

I expect this to be the majority of retain sites in Nokia territory.

FFV4-65C-R3-V1 (DoDeca)

AHFIG radio right behind for G1900+L1900+L2100. (ports 5-8)

[OPTIONAL] Legacy RRUS31 on the other 4 mid-ports (ports 1-4)

Ports 9-12 low-band unpopulated.

Nokia AEHC for 2500.

I have not seen a single retain site with lowband currently, but I have seen them with just 4 of 12 ports populated for 21/19.

Oddly enough, looks like some recent Omaha discussions here.

Been in Omaha for CWS, and have been passively mapping Sprint Keep sites.

I've been mostly inside the bounds of I-480/Dodge on the north edge, I-680 on the west edge, I-80 on the south edge, and Council Bluffs on the east.

Looks like I've picked up ~40 keep sites in just this area so far, and not a ton of different routes taken.

Where I'm staying, roughly I-80 and 72nd, I have L600 on the Pink side, and B25/26 on the Sprint side, and it has been really rough on B25, just 2x 5 MHz B25 carriers.

T-Mobile is great when there is a native macro, and Sprint is great when B41 is available.

I'm sure when these sites get integrated and the spectrum get integrated, its going to be pretty great.

Far ways to come from Rosenblatt in 2007 with no service outdoors, and service only by the window in the Embassy Suites in Old Market.

I'm sure there are tons of keep sites further west/north/south, so here's hoping that Omaha will be a first class city for magenta soon.

EDIT: To add, holy cow the Verizon mmWave build is bonkers here. I can't go anywhere without seeing their small cells on every block. I remember in 2018 or 2019 I saw quite a few LTE small cells downtown, and they've added so many more with mmWave. Absolutely crazy.

3 hours ago, Trip said:

I've seen the T-Mobile TACs change a number of times around here, too.  I actually strongly dislike the T-Mobile TAC setup, which looks like a trainwreck to me, just like their GCIs and PCIs.  I would be willing to go through the trouble of recompiling my entire database for T-Mobile if they would fix their GCIs, PCIs, and TACs to fit some kind of sane and sensible pattern.  I fear that will not be the case.

- Trip

The lack of structured PCIs bothers me to no end, but as I have heard, at least in the Nokia markets, they're just letting the OSS assign PCI based on *science*.

The eNB was ever so slightly structured, until they exhausted their original eNB pool.

Central was eNB 20000-39999, East was 40000-59999, South was 60000-79999, and West was 80000-99999.

Then, especially with Nokia Flexi basebands, you have to start adding 1-3 more eNB, exhausted the eNB pools in those schemas. I know South has a large number of 10000-19999 eNB, as well as 100000+. Oh, and the Nokia Flexi small cells, one eNB for L2100, another eNB for L1900. So 2 eNB burned for each small cell. So it went to hell quickly.

And then with the AirScale baseband swap in, mostly occurring with 600 overlays, they're now able to collapse the LTE layer down into a single eNB, and now a bunch of wasted eNB they are unlikely to recycle (which would be horrific for mapping apps).

The TACs had some semblance of structure at some point, where you could more or less identify the market based on the second and third digits, and fourth and fifth were individual tracking areas within the same engineering market. But of course, they didn't adhere to their own standards.

It's reminiscent of the TMO site ID structure, which will vary based on whether it was Omnipoint, Voicestream, Powertel, Aerial, et al, who all had different site ID structures, and they just merged them all together.

On 3/18/2020 at 5:39 AM, Trip said:

I'm looking at the MLS data for the past few days and it looks like the TACs may have changed in some areas where T-Mobile's TAC values begin with 3 or 4.  For example, in Columbia, SC, it looks like one of T-Mobile's TAC values is 31598.  It looks like any Sprint data collected from that area since at least 3/16 now has a TAC value of 31598, rather than the older 4353 value.

- Trip

I am late to the party, but noticed today that Sprint's TAC in the New Orleans metro was matching T-Mobile's.

T-Mobile has played with the TACs a number of times over the years in the southeast LA markets, and after splitting the New Orleans metro in two, they re-numbered again within 3-5 months, which I thought odd.

Also, being that you mentioned anything with a 3xxxx TAC being possible, the 3xxxx TAC range is used in the "South" region.

1xxxx is the "West" region.

2xxxx is the "East" region.

4xxxx is the "Central" region.

5xxxx are femto.

On 3/18/2020 at 11:51 AM, pokefan02 said:

I can confirm tac has changed in the Louisiana market in Southwest LA

I'm curious if things are changing in Lake Charles, as LC falls under the Houston market. I know the UMTS sites were at 32093, and the GSM sites were at 32247. 

The NOMO (New Orleans/Mobile)/Houston boundary is super weird, because it is one of the few that doesn't follow MTA boundaries, because Alexandria should be part of Houston, and its part of NOMO (after being part of nobody for the longest).

Iowa is a NOMO site, but the two sites north on US-165 are Houston, as is Kinder and sites on US-190 west of Kinder,  but Oberlin and Ragley are NOMO. Its a super weird boundary.

I know NOMO has re-TAC'd all of their UMTS sites to the new TAC schema, so that would be all the way to Pensacola/Destin, and up to about Hattiesburg.

I haven't seen or figured out what they'll do with the GSM (no UMTS) sites yet.

Very odd, especially considering that I still haven't seen my T-Mobile phone roam onto US Cellular.
- Trip

Afaik the only places where TMUS is roaming onto USCC is where USCC has VoLTE live. Which is Iowa and Wisconsin from what I've heard.

On top of that, they're whitelisting IMEIs right now, mostly from Apple and Samsung, but read some were able to get their LG IMEIs whitelisted and roaming started working.

So it's very much an in progress situation.


Sent from my iPhone using Tapatalk

She was able to FaceTime on 256 kbps download?? I'm not familiar with the bandwidth requirements for FaceTime, but is video chat even possible at those capped speeds?

After UMTS was turned on in my market in 2012, GSM was greatly freed up, and I was able to FaceTime over EDGE on an iPad.

 

It was not optimal, but PoC it worked.

 

 

Sent from my iPhone using Tapatalk

I forgot to mention the phone. My mother has a very good working Microsoft Lumia 950xl that I'll be using until the unlocked Sony XZ Premium becomes available, which at that time, I'll give the Lumia back to her.

For reference, as I'm sure you know, that very good working Lumia 950 is very good at not working with VoLTE and wifi Calling on T-Mobile you're going to be using it on.

 

So don't expect the very best service, especially with the voice issues you have had.

 

Just pointing out that you're not doing yourself any favors in setting yourself up for success with that device.

 

 

Sent from my iPhone using Tapatalk

But pretty accurate, at least in my experiences.

YMMV.

 

I've driven a ton of highways and a ton rural areas.

T-Mobile has 100% of sites LTE online in the state of Mississippi. I'm not sure you can point to a more rural area.

 

They beat CSpire in a lot of highway LTE coverage, definitely in interstate LTE coverage, statewide more sq miles than CSpire easily.

 

Is it perfect? Not by a long shot.

Are there dead spots? You bet there are.

Is there band 12 in MS? On 4 sites on I-59, and one site in SW MS. In lay terms, no.

Does it work on "highways and rural areas"? YMMV, but its a non-no answer in my book.

Sometimes I want to try T-Mobile but the whole b12 or edge on towers like on highways and rural areas bothers me.

 

That sounds like an overly pessimistic expectation.

Sprint acquired some AWS spectrum in western Virginia and a 700mhz A block license in Memphis last year. If I were them, I would try to work out a trade with T-Mobile for that 10x10 800 block in Myrtle Beach and/or some PCS spectrum in spectrum constrained markets. The 800 spectrum T-Mobile holds in Myrtle beach is contiguous with Sprint's 800mhz so Sprint could run a 15x15 band 26 carrier (they would need to get equipment recertified) if they were to acquire it.

Don't count on it.

 

700a in Memphis is encumbered by Ch51.

Also, I believe that was a short term lease from CSpire. Still in their hands as far as I can see.

 

And T-Mobile's interest in West Virginia is roughly zero. They might consider a ROB project if they held 700.

 

But with USCC roaming supposedly looming on the horizon, likely no interest at all.

 

Don't get your hopes up.

 

 

Sent from my iPhone using Tapatalk

Can anyone tell me if its true that Tmobile uses Band 5 for LTE in Myrtle Beach NC. Someone on Howard Forums claims they do?

Its true.

 

10x10 L850, plus GSM in 850.

 

 

Sent from my iPhone using Tapatalk

My Tmo DL speeds have been steadily dropping all year on almost all sites.  And now I'm down to several sub 1Mbps sites at peak times in the South Puget Sound.  

 

Assuming near the Olympia area, you have some heavily fragmented spectrum holdings.

 

You've obviously done a ton of mapping on CellMapper, and I only see 1c L2100 @ EARFCN of 2350.

 

Either, they still have U2100 on-air (seriously doubt), or somehow your device never gets pushed over to 2c L2100, which should have an EARFCN of 2200. This would be a drop in doubling of on-air LTE, sans L700 layer.

 

On top of that, you also have fragmentation in the PCS holdings, but its not unreasonable to see them move to 3c 10x10, 2x L21 1xL19, with U19+GSM in the fragmented PCS block, as well as some GSM sprinkled certainly at the guard band of the LTE block, and possibly selective RB sharing with LTE where needed.

 

Very, very odd you haven't seen 2c L21 there. GCI pattern is x65, x66, x67 for 3 sectors of 2c L21, or sectors 101,102,103 in decimal.

You need to start with customer lifetime value.

 

1) Cost to acquire the customer (say $300/customer)

2) Then take churn as a percentage to come up with the number of years you will have a customer (say 1.5% churn/month means that if you had 100 customers you would lose 1.5 of them each month which means to stay even you need to add 1.5 customers each month @ $300/customer to keep your same customer base

3) Come up with a weighted average life of your customer base and the expected cashflows you will get from those customers and take the present value of those future cashflows until the customer is expected to churn out (on an average basis)

4) So then you take the customer acquisition cost + the PV of the value of all the payments you will get based on your churn metrics and that's how you come up with customer lifetime value.

 

So if by offering two free lines to reduce churn, you are potentially making a lot of money by offering this promotion without a big spend on customer acquisition (aka marketing) costs.

 

It's a great promotion, very innovative - even with a $20/sim card cost. In reality, what T-Mobile is doing is giving people free devices with subsidies on the credit side to make sure again they don't churn out. You could walk into the store and T-Mobile will give you two free lines on nice android tablets just to make sure you don't leave for two years.

All while diluting ARPU/ABPU, no?

 

 

Sent from my iPhone using Tapatalk

I hope you all are prepared for more congestion! As we all know, the holiday season is coming and it looks like in Q4 tmobile might report it's biggest growth in company history. I see a few markets already tapping out on spectrum and more markets are opening more bandwith to make up for congestion, but customer growth at this rate will be hard for tmobile to keep up with. If you guys know about anymore movement that T-Mobile is currently making in your market perhaps to lift congestion.. let us know !

 

 

Wide spread doom and gloom is a bit misguided. There's congestion anywhere there is a large mass of people.

 

In my market, they are in the throws of adding L1900 and L700 layers, and there are at least 3 new colocations in the metro in the works currently. Not to mention the current modernization is bringing AWS-3 capabilities with it.

 

 

Sent from my iPhone using Tapatalk

US Cellular doesn't support VoLTE to my knowledge and is a CDMA carrier, so it would be quite unusual.

 

- Trip

US Cellular has been trailing VoLTE for a while now I thought.

 

In Jam 15, they expressed intent to trial in "select cities" within the year.

 

In Jan 16, they expressed plans to launch in one market, following trials in 3 markets. They also talked up the VoLTE roaming benefits.

 

In May 16, they spoke up about the 'VoLTE gap' for CDMA carriers, and didn't include any other time tables other than there was 'no rush' to ship the product.

 

Didn't want to link Fierce three times as sources, but quick googling should yield results corroborating that summary.

 

 

Sent from my iPhone using Tapatalk

True I didn't know that manufacturers other than Ericsson made those type of panels. Which band do you think their for in my market Band 12 or 66?

Bands 2/4/12. There is no AWS-3 spectrum owned by T-Mobile in that market. However, yes the antenna is capable of that frequency range.

 

If Jacksonville follows other markets getting this antenna, you will have L2100+L1900+L700+U1900 operating from this one antenna.

You will have a second antenna for G1900, and likely U2100, since Jacksonville is a 50 MHz AWS-1 market.

 

Keep 1c U2100, and for now 1c U19 for 2c UMTS, get to 10 MHz wide on L1900 from day 1. This gets to 70 MHz deployed LTE. (10 MHz of GSM, 20 MHz of UMTS)

 

Later as traffic migrated from UMTS to LTE, sunset U19 in favor of 15 MHz wide L19. Gets to 80 MHz of LTE (of 100 MHz of usable spectrum), with 1c U21 plus GSM.

 

With a tight enough grid, 80 MHz of LTE should handle T-Mobile's subscriber load for a while. Obviously high traffic sites will be slow no matter what, but sector splits, further densification can tighten up weak points needed as customer growth continues.

 

 

Sent from my iPhone using Tapatalk

It's definitely a T-Mobile site for sure. I found it odd that Ericsson's equipment would be in a Nokia market.

There's a very good chance you are mistaking the new Nokia RAS for an Ericsson AIR.

 

It is very beefy, due to incorporating 2x RRH's inside the antenna enclosure.

 

Typical configuration for non-AWS-3 market would be FHFB + FRBG internal, FRIG external.

 

https://www.reddit.com/r/tmobile/comments/4sff69/brand_spanking_new_nokia_flexi_ras_radio_antenna/

I've been seeing Ericsson AIR panels being installed here in Jacksonville FL. I. Automatically assumed it was for the recently acquired Band 12 spectrum. I guess it could be for Band 12, Band 66, or Both.

I could be wrong, but I feel very confident in saying that any AIR antennas would be those of either ATT or VZW, as Jacksonville is a Nokia infrastructure market.

 

I would imagine since Jacksonville is a late L700 market that T-Mobile may be going straight to NSN RAS, which affords 4T4R on L2100 and L1900, and 2T2R on L700.

 

 

Sent from my iPhone using Tapatalk

I know markets are getting new equipment for B66, namely new Ericsson AIR units. I think marketing may be jumping the gun with saying the spectrum is "on air" which is why it's intentionally vague. There's no way to fact check them because there are no handsets out in the wild that support B66.

Correct,

 

Ericsson is currently seeding B66 AIR units into the wild, likely as a replacement to B4 AIR units, which is why we are likely seeing these in non-AWS-3 holding markets such as LA.

 

Nokia is also seeding their B66 capable RAS units (FASB) in many markets across the country, mostly those that were late L700 bloomers.

 

Beyond the RAS addition, the existing FRIG radio just needs to be swapped for a FRIJ that is capable of AWS-3.

 

 

Sent from my iPhone using Tapatalk

T-Mobile announced that they have B66 on air. The question is where? Extremely vague press release.

 

I know of at least 1 engineering market gearing up for immediate AWS-3 rollout.

 

It's an antenna swap, along with a radio swap.

 

The nice part is that this antenna and radio swap is pretty much happening at the same time, thanks to L700 rollout.

 

Soon...

 

 

Sent from my iPhone using Tapatalk

What happens is that 90% of the customers just accept it and keep on moving, and the 10% port out.

 

Risk vs Reward, and it won't stop their subscriber gains.

As someone affected, I don't welcome it with arms wide open, however, I'm not porting out.

 

Of about 29 months on my rate plan that includes 5GB of tethering, I've used 9.2 GB of that over that lifetime, 1.5GB of which was last month.

 

So I average ~250-300MB of tethering monthly, mostly SSH and email traffic when I'm away from my desk. For me, deprioritization should not affect my shell sessions, and if it does, then that becomes an issue, but in my use case, its a non-issue to me personally.

 

And if it does become an issue, I have a company issued Verizon hotspot to take my business to.

 

 

Sent from my iPhone using Tapatalk

The whole point of B26 is to only serve customers not reached by B25/41.

I've had this discussion with others before, but I've always been of the mindset, especially in urban areas, that B26 should have been viewed as more of a capacity play.

 

If you are in ~2014 Sprint land with the majority of the network with a single G block B25 carrier, and B26 carrier, you have 50 RB available for traffic.

 

However, you are limiting B26 traffic to basically QPSK frames to the UE, because you are only getting to B26 by way of running out of rope on B25 in terms of signal. Never mind B25 was out of rope on capacity in many cases, compared to B26.

 

Why not serve more bits up, increase total served throughput at the site and load balance across B25/B26 to get more 64QAM frames out of B26? 1x800 takes care of CS traffic at cell edge as is, and data on B26 wasn't going to be great with low RSRP/SNR to begin with.

 

That was always my take on it at least. I get the argument for creating a "consistent experience" but it sounded like it was a consistently inconsistent experience as a result.

 

 

Sent from my iPhone using Tapatalk

I travel a lot so want to make sure coverage and speed is reliable

Travel the entire country? Only a region of the country? Travel by air/bus/train/car?

 

Where is home base?

 

Any type of starting point could steer in you in one direction or the other.

 

 

Sent from my iPhone using Tapatalk

That's really odd. Head scratcher.

 

T-Mobile has a TON of spectrum in Jacksonville, with 50 MHz of AWS-1 and 20 MHz of A block and 30 MHz of C block.

 

This 10 MHz carrier is using up the full A block.

 

In my mind it makes far more sense to deploy in the C block, even if just at 10 MHz, because you have the runway to get to 15 MHz wide.

 

They could easily have 1c U2100, at least 1c U1900, 5 MHz of GSM, 20 MHz L2100, 15 MHz L1900, and 5 MHz L700.

 

That's a conservative approach. Considering that they've been moving quickly to 1 UMTS carrier in lots of markets.

 

 

Sent from my iPhone using Tapatalk

For GSM, here's what I saw this morning:

 

687

688

689

709

711

712

713

714

734

735

 

- Trip

 

With EARFCN of 1025, you have a 5 MHz carrier, centered at 1972.5, extending from 1970-1975.

 

GSM channels, being 200 kHz wide, ARFCN 711 is centered at 1970 exactly, with 712 at 1970.2, 713 at 1970.4, 714 at 1970.6, 734 at 1974.6, and 735 at 1974.8. So we clearly have overlapping GSM with LTE, likely by way of selective resource block shutdown.

 

Oddly, UARFCN of 587 is centered at 1967.5, and Ericsson generally can reduce UMTS bandwidth to 4.2 MHz wide, which would be 1965.4-1969.6, though I have heard that Ericsson has recently matched Nokia in being able to achieve stable 3.8 MHz wide UMTS carriers. The reason this must be a 3.8 MHz carrier is because at both bookends, you have GSM carriers, at 688 (and 689) and 709 and 1965.4 and 1969.6 precisely.

 

So assuming 3.8 MHz wide U1900 carrier, 689 is still right at the end of the carrier, so something strange is definitely going on.

 

Either way, they're stuffing GSM into the LTE carrier (it looks like), which is pretty crazy and exciting.