Conan Kudo

Blackjack dealer

Tennessee Williams

Well they've also been wrong the last two quarters too. They always forecast Sprint to lose more than they actually do.

To their credit, they are usually rather close on the number of handset line losses, though.

Which makes for interesting builds on the T-Mobile side. Are they looking to upgrade their base stations or keep their existing builds throughout the existing network? I read that T-Mo will upgrade their EDGE network using ground-mount RRUs, but what about their existing LTE/HSPA?

The HSPA+/LTE network uses tower-top electronics. And any cell sites getting 700MHz will use tower-top electronics too. That means when those TMAs are removed at the antenna, full multi-band 700/AWS/PCS RRHs go up in their place. The PCS-only ones aren't getting it for now because T-Mobile doesn't want to go up and down to redo the antennas several times in succession. It would be a waste of time and money to do that.

So the GMOs are a stopgap until they know for sure about 700MHz in the PCS-only zones and the supply for tri-band RRHs improves, at which point they'll convert to tower-top electronics using the RRH up at the panel (or AIR systems in Ericsson areas).

Essentially, the upgrade has been split into two stages to ensure it gets done faster overall. After all, T-Mobile still has to get that AWS deployed at the very least, too.

I've been having a great sidebar conversation with Neal on Google Hangouts. He has corroborated that Tmo also suffers with Ericsson significantly. He also noted that Tmo took the entire Southeast US from Ericsson because of issues. So Ericsson is the common denominator here. Not Sprint.

 

Neal speculates that Ericsson keeps AT&T happy at everyone else's expense. And there may be some truth to that.

 

Robert via Nexus 5 using Tapatalk

Since I've been sprinkled into this conversation, I figured I might as well jump in anyway.

Yes, the portion of the T-Mobile US network for the Southeast originally used Ericsson. This is because the Powertel network was the basis for it. By 2000 (when Powertel was acquired by Deutsche Telekom along with VoiceStream), Powertel had PCS licenses throughout Mississippi, Tennessee, Alabama, Georgia, and Florida. Throughout that region, Powertel used Ericsson's GSM system to power its PCS network. For reference, SunCom used Nortel, Omnipoint used Ericsson, and Aerial used Nokia. VoiceStream used Nortel and Ericsson in select coastal markets, and Nokia everywhere else. The Nokia relationship extended to the Cook Inlet/VoiceStream JV and the Iowa Wireless Services JV.

In 2008, T-Mobile started rolling out UMTS using Nokia, Lucent, and Ericsson. Lucent was kicked out sometime later, and NSN+Ericsson remained. By 2012, though, T-Mobile had removed Ericsson from the Southeast. In fact, I had personally witnessed several sites in the Southeast where Ericsson was removed for Nokia and I've been told that it took place all over the region as part of the 3G/4G upgrade and it was happening in other regions too.

By now, the only places where Ericsson is in use are the Northeast Corridor and California. In un-upgraded (2G only) territories, Ericsson is still present in the Midwest, portions of the Southwest and non-California West Coast region. In these areas, the BTSes and the TMAs are coming down and are being replaced with new ones from a different vendor.

Sprint will be the first carrier (via Boost Mobile) to offer the iPhone 6 and 6 Plus in the prepaid market:  http://www.bidnessetc.com/26143-boost-mobile-to-launch-iphone-6-6-plus/

Sprint's Boost/Virgin also don't have BYOD capabilities like MetroPCS, TracFone, and other GSM prepaid brands. Since most CDMA prepaid brands aren't allowed to activate iPhones, this is notable. What is odd, though, is that Virgin Mobile isn't part of this. 

Honestly, I seriously doubt TWC could be threatened by Google Fiber in NYC. I live here, and know just how completely entrenched TWC is with the boroughs. If Verizon can't even get into buildings, what makes you think Google will be able to?

Most people like Google. Most don't like Verizon.

Iusacell used to be a CDMA carrier. Did they actually rip out the CDMA network or do they maintain both?

The CDMA network is gone, as of the end of the second quarter of 2014. All the information about the CDMA network and the CDMA->GSM transition pages are gone. The Unefon brand (which was used for CDMA in the past) has been relaunched as a GSM/UMTS operator. Iusacell is planning on launching LTE in the first quarter of 2015 now.

It doesn't affect Verizon, since they switched Iusacell roaming to GSM/UMTS back in the fourth quarter of last year, like they did with Telus at the beginning of last year.

SoftBank purchase.  Not Sprint purchase.  Other than having the same parent company, this will not impact Sprint directly.  Masa is doing other deals and will get his own Marcelo for Mexico.

It's much more beneficial for SoftBank because Iusacell is a GSM/UMTS/LTE operator. If SoftBank takes a stake in it, then it can synchronize and expand procurement operations for network gear and devices between the two companies much more than it can with Sprint and itself.

Yeah but NYC isn't like podunk suburbs that Google can micro-trench and run cables. 

That's beside the point. Google Fiber is a threat to TWC there. Also, if TWC/CMCSA is killed, then TWC will be incentivized to move faster on DOCSIS to compete against the specter of Google Fiber. TWC has lost massive share in Kansas City in Google Fiberhoods, and they don't want to see that happen in NYC.

Comcast has the same fear in some of the places they operate against new fiber providers (C Spire, etc.).

Great point on different modernization efforts.  I love the fact that Sprint basically forced T-Mobile's hand into upgrading rural areas and in many ways, T-Mobile has forced Sprint to refocus on deploying B41 in urban areas.

 

In terms of CAPEX, when we look at the two (not adjusting for subs, clearwire, etc) we see that both have invested roughly the same in their network.

 

Sprint 2010-2014

21,895

T-Mobile 2010-214

19,770

 

As far as future capital spending, both are expected to continue to invest heavily.

 

While sprint has already spent a lot of money, they are also much more leveraged (net debt/ebitda=468.5% vs. 311.3% for T-Mobile) .  While we can discuss the advantages/disadvantages to each business strategy, that's a different topic.  You mention that, for example, T-Mobile doesn't have low band spectrum yet.  You can toss in T-Mobile acquiring a nice chunk of 600MHz for a few billion and still have less debt than sprint.  

 

They are different companies with different purposes with different strategies and business plans.  

Another key point to keep in mind is that T-Mobile's CAPEX is not expected to go down within the next 18 months. If it stays flat (as they've been saying for the last couple quarters), then it is effectively a capex raise compared to the business plan that T-Mobile pitched back in 2012. And T-Mobile is gradually growing its 700MHz footprint in order to have an alternative to the increasingly dicey 600MHz auction. Sprint itself underwent a massive 10 year program (that is just now nearing completion) to make its ESMR spectrum usable for cellular services.

And T-Mobile is periodically selling bonds to refinance its debt and de-lever faster than it would organically. Of course, most people don't focus on this aspect of the bond raising. Sprint does this too (though much less often), and it is foolish not to do it, especially since both are not considered junk stock (meaning selling bonds isn't usually difficult).

As both companies grow (yes, I do mean grow; I want to see Sprint eat Verizon's lunch!), we'll see the financials improve as the cost is spread across more customers and absorbed by higher total revenues.

Which I seriously doubt will make it here.

Well, there are rumors that Google has been negotiating with NYC about adding it to the Google Fiber expansion, but we'll have to wait and see if anything comes out of that.

Look at NYC, where Verizon is contractually obligated to deploy FiOS, yet can not get their foot in the door with a ton of apartment buildings. My own block has FiOS, but my building doesn't have access.

Yeah, I know, but FiOS and U-verse territories are unusual. They existed before the Gigabit fiber thing, and their motives are different (replacing existing copper lines). I'm talking about fiber deployments like EPB, Google Fiber, C Spire Fiber, et al, who exist for the purpose of being actually competitive.

The Doctor

General Practitioner

T-Mobile just signed on their first LTE roaming partner: GCI in Alaska.

 

http://www.lightreading.com/mobile/4g-lte/t-mobile-gci-ink-lte-roaming-deal/d/d-id/710962?

 

Looks interesting. I wonder if it is through the roaming hub or standalone. I see they also run a CDMA network, but I imagine that is either on its way out or skeleton roaming for Sprint and Verizon?

The CDMA network used to be bigger (ACS ran the network before merging its spectrum and towers with GCI's GSM/UMTS/LTE network to form the AWN joint venture), but most of it has been turned off to make room for UMTS and LTE. More of the ACS CDMA network is being turned off all the time, since ACS is now selling GSM/UMTS/LTE devices.

The AWN CDMA network is basically a skeleton network that operates in a portion of ACS' wireline footprint. The other CDMA network in the state belongs to MTA (Matanuska Telephone Association), but only covers small rural areas. In fact, MTA has halted CDMA spending to focus on Verizon LTEiRA 4G deployment. MTA will also discontinue selling CDMA wireless service to customers in favor of using the Verizon Wireless brand with LTE-only service.

Agree with the statement but disagree that docsis 3.1 modems will be in the retail channel at any point next year.  I hope they are, cause I'll definitely want one!!! (but given the history of cable modem launches, I can't see that happening, hope I am wrong)

I think the difference this time is the threat of fiber internet service in several big MSO territories. I suspect that will speed things up considerably.

Well, in the case of TWC, they are definitely going with 16 channels in new maxx service areas.  A friend was recently received the upgrade to MAXX and went from 8 channels to 16.

 

Yeah, OFDM is going to be great, but you need to wait until TWC upgrades the CMTS to support the wider channels.  It will be great when they do, but again, it's like people that were saying "oh wait for DOCSIS 3.0, it's about to come out" and that still took years.  I guess a lot has to do with where you are at in your "modem upgrade cycle".  I am on a old 4 channel docsis 3.0 modem and skipped the 8 channel modem.  I'll do the 16 channel 3.0 and probably get the second gen 3.1 modem when my local CMTS is fully upgraded to support it.

 

There are some really really good deals on cheaper 8 channel highly rated cable modems that you can get nowadays, so maybe a 8x channel for ~50 bucks while waiting for a 3.1 might not be a bad way to go.

It's quite likely that DOCSIS 3.1 cable modems will also support DOCSIS 3.0 with 16 downstream channels and 8 upstream channels. So for that reason, waiting for DOCSIS 3.1 modems to roll out next year makes sense, because you'll still gain the advantages of a DOCSIS 3.0 16x8 cable modem, you'll just be more future proof.

Right, like DOCSIS 3.0 16x8 modems that came out last year.

 

Docsis 3.1 24x modems will probably be out in 2016.

Realistically, eight channels is likely to be the maximum that any DOCSIS provider will set up per user downlink on cable. And like smartphones, cable modems are driven by MSO requests. Very few (if any) DOCSIS providers requested modems with 16x8, so only one actually exists (Arris' modem).

DOCSIS 3.1 will be different because it will pack more bandwidth into the same number of channels (or use fewer channels to provide the same bandwidth). And the spec was finalized earlier this year. It makes complete sense for DOCSIS 3.1 compatible cable modems to come out next year, because they know there will be demand for it.

After years of suckling the Verizon FiOS teat, I unfortunately moved to an area which is served by Time Warner Cable.

 

Thankfully they have upgraded most of their network, and are giving free speed upgrades like candy. I'm paying for "Turbo", which is 20 down/2 up, and getting close to 60 down/10 up, with my own Netgear N450 modem+wifi.

 

Unfortunately there is a lot of wifi interference in my building, and I'm looking to upgrade my modem with a dual band option, 5ghz/2.4ghz. I am not opposed to replacing the N450 with a standard modem and an external access point, but am trying to save money.

 

What modem + WIFI or modem + AP would you recommend? 

Personally, I would recommend buying the T-Mobile ASUS TM-AC1900 router. It's basically an RT-AC68U for half the price. They'll sell it to anyone (just say you're a prepaid customer if they ask). I'll vouch for the RT-AC68U (which I currently use) as a totally fantastic router worth having. If you can swing the 25% accessories discount coupon (CLASSACC25) through T-Mobile telesales (sometimes you can, sometimes you can't), then you can get the router for $75 instead of $99. In either case, it's a bargain and definitely worth getting.

I've stopped trusting Motorola modems since Arris took over. They've not been as reliable for me. I swapped out my Motorola SB6141 for a uBee DDM3521 at my home. My brother has had his Motorola SB6180 swapped out for a ZyXEL CDA30360 in his college apartment. We're now getting much better experiences on our broadband connections. I see that Time Warner (like Metrocast in my brother's college town) doesn't appear to support the uBee like Comcast does where I live now. So the ZyXEL (which is supported by Time Warner) would be a good choice.

The ZyXEL modem+TM-AC1900 runs you ~$175 if you can't get the coupon to work. If you can get the coupon to work, it's only ~$155. The uBee modem+TM-AC1900 runs you $159 without the coupon. With the coupon, it's only $139. 

With new DOCSIS 3.1 cable modems coming out next year, it's worth holding onto your money for now and not spending it on a 16x4 Arris SB6183 that you're not going to get your money's worth out of, anyway. 

Cement blocks

Claro seems to match up with Sprint PR roaming coverage. And Sprint now has EVDO roaming in Alaska again. Anchorage, Prudhoe Bay, Soldotna, Kenai, Juneau, etc. have 3G again from Alaska Communications!

I thought Sprint roamed on Open Mobile? Claro hasn't been a CDMA/EvDO network in years (América Móvil converted it to GSM/UMTS after acquiring the network from Verizon in 2006).

General subscriber access to these customized, on premises Wi-Fi routers has to be the T-Mobile end game.  Legere is not the altruist that many believe him to be -- he and the pinkish executive team have plenty of ulterior motives up their sleeves.

 

That said, I am intrigued.  If T-Mobile can sort out the security and bandwidth management issues for general sub access, this could make for an interesting "small cell" deployment.

 

Better yet, partner with or buy out the Globalstar TLPS spectrum for unencumbered 2.4 GHz Wi-Fi.

 

AJ

You know as well as I do that waiting for solving the security problem isn't altruistic, it's simply good business sense. Especially in this day and age, people are increasingly aware of security problems and taking companies to task about it. And in an competitive environment, no one can afford to screw up like that. Of course, Comcast isn't in a competitive environment, so they don't care.

I do think there are some interesting opportunities there. I didn't think about the Globalstar 2.4GHz spectrum, but you're right that it would be an interesting way to go about Wi-Fi powered small cells.

For its Free Mobile subsidiary, Iliad puts customized Wi-Fi routers on its subscribers' premises in France to expand and fill in its wireless network for all subs to use.  Hmm, coincidence?  Is no one talking/writing about this potential connection?

 

AJ

That was specifically asked at the event. Legere said that they aren't doing it because of the security issues regarding it, which is partly why there's so much backlash at Comcast for doing it (the other part being that Comcast is doing it without asking subscribers for permission first). He said that if T-Mobile can figure out a way to securely offer such capability, they may do it. But for now, they won't.

Band 4 over the new AWS-1+3 band wouldn't cause any significant change. The new band class would be specified by the end of the first quarter of 2015. As a superset of band 4, supporting both will not require more parts that what is already required to support band 4, just an expanded duplexer and wider range PA to support both band classes.

And AWS-3 spectrum would have been very useful, no matter what. It lines up nicely with Sprint's current site layout, and it would have provided a nice capacity boost with next-generation multi-band phones.