reedacus25

Looks like T-Mobile is preparing to light up LTE on PCS here. I found a single site in my log from Alexandria which has a 08 and 09 sector, and the EARFCN was 1025. That corresponds to the upper half of the T-Mobile 10x10 in PCS.

 

I also don't seem to have HSPA anymore, though I'm not sure I ever had it on the DAS at work before. I'll see if it's still around when I leave today.

 

EDIT: I eventually got my phone onto both HSPA and GSM for T-Mobile. So it's still running after all.

 

- Trip

Sectors 07/08/09 sound like the L1900 found in Los Angeles, so I assume this to be an Ericsson market.

From what little I hear, T-Mobile is getting pretty aggressive with selective resource block shutdown to accommodate an 'in-band' GSM channel within the LTE carrier.

Would be very interesting to see what E/U/ARFCN's you could pull out of it. Especially given the limited PCS holdings in that market.

That would also put T-Mobile at 60 out of 70 MHz usable spectrum devoted to LTE, which is pretty insane.

Sent from my iPhone using Tapatalk

In a typical Sprint market, band 41 utilization almost certainly will be greater than 256-QAM utilization. Band 41 may have a "short umbrella" in the macro environment compared to that of band 25 and band 26. However, the 256-QAM umbrella will be even shorter, significantly so.

 

AJ

Without a doubt, B41 util > 256QAM framing in a traditional macro grid.

However the principal at a zoomed out view holds true in that there is a ton of capacity closer to the center of the cell, albeit B41 has a much larger umbrella, and a more gradual taper with 64QAM->16QAM->QPSK across 20-60 MHz versus 256QAM->64QAM small radius across 5-20 MHz. Maybe more with CA.

Sent from my iPhone using Tapatalk

Yes. And people who are eating up this T-Mobile announcement are buying into smoke and mirrors. Both 256-QAM and 64-QAM utilization almost certainly will be under 10 percent of the time.

 

You know, people criticize Sprint because they are not on band 41 often enough to their liking. Well, T-Mobile users are going to be utilizing 256-QAM and 64-QAM far less often than that. However, adaptive modulation lurks underneath the hood, while band selection is more visible to end users. So, people will be oblivious to the fact they are rarely, if ever actually using 256-QAM and 64-QAM. They just see the announcement of higher peak speeds and think, "Ooh, new, shiny."

 

AJ

My initial thought to this was the comparison to the B41 short umbrella at the macro layer compared to B25/B26.

You get much higher capacity (b/Hz) compared to edge of cell.

Sent from my iPhone using Tapatalk

I had Edge before crossing the river into Baton Rouge from the west but between the Texas border to Baton Rouge, there is a lot of area that is very badly covered. I had some massive problems around the Hollywood Casino in BR.

 

Denham Springs was pretty bad as soon you drove into sub-devisions and I got always NO SERVICE, no fall back roaming.

Now I'm also curious what device this was, because my daily driver an iPhone 6, even without B12, still has almost steady LTE service except in my stated problem areas. And while T-Mobile isn't blanketing Denham top to bottom, T-Mobile is colocated with Sprint on 2 sites, and T-Mobile has a third site in Denham proper, and so most anywhere LTE coverage should tilt in favor of T-Mobile, though Sprint will edge out T-Mobile with 1x, especially 1x800 outside of LTE.

Here's all the CellMapper data from myself and others. Personally I've used Samsung Galaxy Avant and Moto E to map to CellMapper, just to add a level of transparency for what devices I personally used to map in these areas. It pretty well matches my experience.

https://www.cellmapper.net/map?MCC=310&MNC=260&type=LTE&latitude=30.208992405679787&longitude=-92.27828216552733&zoom=9&showTowers=true&clusterEnabled=true&tilesEnabled=true&heatMapEnabled=false&showOrphans=false&showFrequencyOnly=false&showBandwidthOnly=false&DateFilterType=FirstSeen&showHex=false&bands=0,2,4,12

So I took a business trip from Houston to Baton Rouge in I-10 and T-Mobile was not up to the task. My Verizon iPhone with Unlimited data was working without any issue 99.9% of the time while T-Mobile dropped down to EDGE several times. Even in Baton Rouge, LTE data on T-Mobile was extremly slow and not working at times. Ran several speedtests and T-Mobile was at or around 1MB down. Sprint was working well, not as well as Verizon but close and even into Denham Springs, you were able to see what a difference Sprint made in a year. B26 all over the place and areas that had 3G last year are now running B25/26 & B41.

I'm really curious where you had issues along I-10, at least from the LA state line to Baton Rouge.

It gets really thin near Jennings, and the site on the Basin Bridge near Grosse Tete is still GSM only, but I do not drop LTE between TX/LA line and Jennings, and from the drop in Jennings to Grosse Tete.

Also curious where you were seeing issues with TMo in BR/Denham, because even though its only 10x10 L2100, they've got tons of 5x5 L1900 on air, especially along I-10,I-12,I-110, and quite a few sites along/near Airline and Florida. No L700 on air yet however.

My only trouble issues in BR are near I-10/Seigen and downtown BR north of Florida, LSU's campus (and Tigerland). Just curious of the discrepancy in our two experiences. Not saying its perfect, far from it, but we've had two very different experiences.

MAC filtering should not be used for authentication. Carnegie Mellon's wifi network used (maybe still uses) it as the only auth, which blows my mind for a university that's the home of CERT. Before I was a student there and needed wifi, I just cloned the MAC of a user I saw on the network and viola, I had access.

 

Not saying use that as the only authentication method, there are plenty of further ways to restrict authentication to valid devices only, but that's an easy first step to weed out unwanted devices.

Sent from my iPhone using Tapatalk

I'm surprised more companies do not have dual SSID networks, even authenticated ones. We recently upgraded our network here and went completely wireless with the APs broadcasting a secure network which only authenticated on corporate assets, and an authorized guest network which required credentials but no device restriction.

It's not that hard to whitelist MACs for a single SSID, even on basic consumer gear...

Stick the guest SSID on a completely isolated VLAN from the internal traffic, enable isolation for all users on that SSID so they can't talk to other wireless guests, and call it a day.

Sent from my iPhone using Tapatalk

They could've been doing great in Chicago if they had started investing in it sooner, but they do run a pretty lean ship and didn't wanna spend the money a year ago when they should have started. I suspect they held off until they knew they would have 700A in their arsenal since the competition is fierce from the other three, and without it probably would've spent a ton of money and not quite gotten the results they want.

I'll entertain the notion that they likely have been running a lean ship out of necessity, more so than want. So by the time that the money came in to get the ball rolling on projects, it was likely "too late" to head the congestion off at the pass if you will. And here we are. Not saying this is fact, but a plausible scenario.

Sent from my iPhone using Tapatalk

I am in el paso tx and we need band 2 really bad

 

Sent from my SM-G928P using Tapatalk

El Paso has:

• AWS D (5x5)
• AWS E (5x5)
• AWS F (10x10)
• PCS A (15x15)
• PCS B3 (5x5)

AWS D-F all contiguous, all LTE I imagine. So 20x20 L2100 @ EARFCN 2300.

They should also have at least, 2c UMTS in PCS A, pick whatever UARFCN's you want, with the rest GSM.

Right now they can likely support 5x5 L1900 + 2c UMTS (DC-HSPA) at full 5 MHz wide channels.

Coupled with L700, this gets you to 60 MHz of deployed LTE (20 MHz deployed UMTS, 10 MHz deployed GSM).

However, I can think of at least one scenario where I believe they could achieve 10x10 L1900 with 2c UMTS (not DC-HSPA), and GSM. Shove 10x10 L1900 to whatever edge of the A block, and in the remaining 5x5 block, you squeeze a 3.8 MHz carrier, surrounded by GSM carriers. Do the same thing with the B block license, with a single 3.8 MHz carrier, and GSM carriers surrounding it.

Should be enough for a k=4 reuse pattern.

However, I think you will likely see them move to 1c UMTS at full width, and expand the L1900 carrier to 10x10 when UMTS traffic warrants its not being needed anymore. The radios and antennas likely would not have enough ports to support that many separate connections in random spectrum blocks unfortunately.

Hopefully that answers your question. There's a path to it, but it will likely be a gradual transition. And likely around 2018, although maybe sooner, you may see them remove UMTS completely and move to complete LTE in the A block, leaving 5 MHz of GSM for circuit voice, which should be plenty to carry the voice traffic after more LTE device penetration.

The good part about the L1900 deployment in most markets is that it requires little if any tower top work to achieve. A system module installed at the tower bottom, and either a reconfigured radio, or a reusing of existing U2100 lines gets it on air without needing to climb the tower. Therefore, the turnaround time to bringing it online is relatively short compared to say the L700 overlay.

Wow! Band 2 deployment is very speedy just like band 12. I have seen band 2 lte pop up in many places. Is it true t mobile is starting band 2 with 5mhz ?

 

Sent from my SM-G928P using Tapatalk

Depends on a few different variables, some markets are going 10x10 (or 15x15 in Dallas iirc) when there is excess spectrum to devote to it, usually by having 50 MHz of AWS-1 to keep 1c UMTS in 2100, this freeing up PCS for GSM+ LTE.

Sent from my iPhone using Tapatalk

Just thought I'd throw this out for folks looking for roaming LTE. I'm visiting McComb, Mississippi and noticed I'm getting LTE on Cspire. 

 

http://cloud.tapatalk.com/s/577caddc3456f/Screenshot_20160706-020447.png?

 

Not sure if you have debug screen on your device, but curious what the GCI pattern is for CSpire. I know they have B12 in the wild, and I think some B5 in some territories as well. Not sure if they have and B4 deployed yet, but I know in some markets they have gotten to 10 MHz wide on B25, and maybe even 15 MHz wide in Hattiesburg I think was the market they were trying to achieve that in.

Anyways, I picked up CSpire on my Moto E near Natchez, and that device was had B12 disabled in the firmware, so it may have been B5.

Yup. Some don't get it and ramble on about disrupting their sense of Aesthetics.

 

While aesthetics may be a subjective matter of taste, it's hard to argue with wasteful construction to duplicate a utility pole when a perfectly usable existing utility structure is less than 5 yards away.

It grinds my gears to drive through north Florida and see 4-5 monopole structures lined up in a span of 100 yards because no one can build a tall enough structure to support all 4 carriers. It's wasteful.

@greenbastard's argument is actually for small cells, when they are placed on existing structures with minimal disruption to the surrounding areas.

What he doesn't want is a pole placed in the easement, especially in his easement, that sticks out like a sore thumb because it stands 2-3x as tall as all adjacent structures in the area. It's obnoxious, and again, wasteful. I don't think many would argue the fact that there are an innumerable amount of existing utility poles that could handle the job.

Sent from my iPhone using Tapatalk

B12 can't help where I live since I'm in one of the few places where T-Mobile doesn't have it.

 

I'm in Mississippi enough to feel your pain on this.

Sent from my iPhone using Tapatalk

I've noticed Tmo adding B12 to the B2 GMO sites. Which has been nice to expand rural coverage. But the big network issue is the declining speeds. At high use sites, I see the decline every month. Most of the really high speed sites seem unaffected, though. Tmo will need to densify in my area.

In the areas they have had the L700A license in Louisiana (not recent areas like New Orleans/Baton Rouge), they have very much targeted major vehicular corridors.

I-10 from Texas to almost Baton Rouge is B12 the entire way except for a lapse in Crowley.

I-20 from Minden to the end of the license in Delhi, aside from the site in Ruston.

And I-49 from Shreveport to north of Alexandria was all B2 and has been overlaid with B12 now.

Point being, major corridors are being overlaid pretty well in markets where the license has been held.

New Orleans is just now getting to the permit side of things with the L700 overlay after having the license since last fall.

However, if recent events should mean anything, expect some aggressive PCS refarming. New Orleans and Baton Rouge down to single U1900 carriers to make way for L1900. Very interesting to see them break their 2c UMTS soft limit they've held fast to nearly nationwide.

Sent from my iPhone using Tapatalk

I will admit, in the interest of full disclosure, that I'm thinking of porting my currently unused AT&T number into Magenta solely to claim the stock share.

Let me know if you need a referral [emoji6]

Sent from my iPhone using Tapatalk

-- Sent from my Palm Pre3 using Forums

Holy cow. I haven't seen a WebOS device in so long, let alone a Pre3. Mine should still be in a drawer somewhere. How's it holding up after this many years? Was always concerned about its durability.

Sent from my iPhone using Tapatalk

The biggest question mark will be how they choose to deploy the spectrum. Does L700 go on every site, or every other site, or every 4th site? That's the million dollar question.

 

In rural and suburban markets, deploying L700 at full power helps blanket the area with coverage, but does nothing for capacity.

10 MHz of deployed LTE does nothing for capacity? B25 is a 10 MHz carrier. Does 2c B25 not matter because it's just 10 MHz?

While having the ability to cover more POP is great, but to be honest, I see very little merit in deploying 5x5 700 in a market that is completely capacity strained.

Spectrum/capacity constrained + spectrum/capacity = little merit?

Sure it's not going to help post 900 Mb speed tests, but it's going to add 33% more capacity to the LTE RAN, and greenfield at that, without cannibalizing PCS yet. So calling 5x5 not good for capacity is without merit.

Sent from my iPhone using Tapatalk

Not sure what this would do going from 4 to 3 major carriers. But myself I am all for a fifth major carrier. I'd love to see Dish+Comcast or Dish+Charter do an MVNO. Do regular popular MVNO pricing like say 3GB of data for $30/mo but then have it where when you ride Dish's airwaves it doesn't count against your data bucket. So you get VZW LTE coverage through the MVNO that gets unlimited data when riding Dish LTE. And just have Dish deploy their spectrum in the major pops.

So wait is some of the coverage that 700 covers it's just that and edge with no fallback to HSPA?

 

Almost all new coverage that I have seen or come across depends on its location.

New densification sites in urban areas are L2100+U1900 for capacity.

New sites outside that in roam overbuild areas are either L2100+U2100 or L1900+U1900. Both is these irrespective of L700 which is dependent on if it's available.

Then there are L700 only sites. No GSM, no UMTS.

El paso has band 4 + band 12 CA. It did reduce congestion by a lot. How long will this hold up until they have to refarm the pcs band 2

 

Given their strong spectrum position, it would not be out of the question to see some aggressive L1900 overlay in El Paso.

It would be pretty trivial to achieve 2x U1900 carriers, 10x10 L1900, and keep GSM afloat. Only drawback is having to dismember DC-HSPA in U1900. But I wouldn't be very surprised if El Paso gets some L1900 love due to their ample spectrum, and their known congestion issues in that market.

One thing you have to keep in mind when Sprint is compared to T-Mobile is that T-Mobile's was mainly an overlay. Starting from scratch (installing fiber backhaul) is expensive and time consuming.

In the rural areas, it's absolutely an overlay.

They bring in Nokia system modules, and a single radio that handles GSM and LTE, cut over the feeders, and you're online with GSM.

If you have Ethernet backhaul, you fire up LTE, or you then wait for that backhaul to arrive and then flip the switch. So yes, the rural L1900 overlay is a very efficient overlay project.

However their urban L2100 projects were much larger scope and required almost a full rip and replace. All new antennas, RRH, and TMAs at the tower top, and new converged gear at the base. Backhaul almost always already in place.

But the modernization was not a bolt-on overlay, akin to the now L700 overlay.

iPhone, but that doesn't have a good way of checking what band one is on with Field Test more or less borked at this stage of iOS and Apple's policy on signal apps which is ancient and needs overhaul. Given Apple is good with unlocked phones, I don't see why they can't be better here. I've considered writing Phil Schiller and seeing if something could be done but I don't anticipate a response on his end. Carriers should all be more transparent here as well.

Could be wrong but I think the SE has a 'working' field test screen. i6S/+ still are borked.

It doesn't help that people in Mississippi love their C-Spire.

CSpire has both a lock on their customers, as well as a great network to compete with.

And Sprint doesn't exactly have a great incentive to do anything when they have CSpire to roam onto.

Do you think T-Mobile is doing the PCS refarm in St. Louis yet? I'm going to look for PCS LTE next time I'm up.

 

Sent from my SM-N920V using Tapatalk

same spectrum configuration as Chicago, same vendor, I would be very surprised if the approach weren't shockingly similar to that of Chicago.

Noticed today on the Tmo sub that B2 LTE is now popping up in places, which makes me wonder just how much PCS they will refarm. Right now it's only being used for EDGE and GSM, right?

GSM and EDGE are for all intents the same thing in this discussion.

T-Mobile's modern approach has been to shoehorn all the AWS to LTE, and all of the PCS to UMTS and GSM.

So in your post, the UMTS layer(s) were excluded.

T-Mobile has been very steadfast in their desire to maintain at minimum 2 UMTS carriers in each market. So in this Chicago example, since Nokia is the market vendor, they have 3.8 MHz wide UMTS carriers in the tool chest, and applied that to both UMTS carriers. Leaves 11 GSM channels, which is an aggressive k=3 reuse factor.