Conan Kudo

On the flip side, the same thing was said of Commodore International, Palm, Northern Telecom (Nortel), WorldCom, and many others. They all eventually failed.

Failure conditions

It'll be done by the end of the second quarter next year. Possibly sooner if they keep reaching targets ahead of schedule.

Actually, 700MHz realignment (if we follow US 700 band plan) could eliminate DTV 51 issues. The APT 700MHz band plan has a guard built in at 698-703 MHz (5MHz) which we know is suitable for DTV/Cellular service protection. By rebanding US Lower 700MHz paired blocks to 5MHz FDD each, we can create a 3MHz guard against DTV 51. That would permit 3MHz LTE carriers in A block during the period in which DTV 51 is still in operation (implicitly expanding the guard to 5MHz on uplink by not using the lower 2MHz of the A block). A 3MHz FDD carrier is better than no carrier at all. If we wanted to be really radical to maximize spectral efficiency and open up more usable blocks on 700MHz for use and auction, we could reband to the APT band plan. That would be difficult and require a broad range of supporters in the industry, though.

Alarm clock

Are you even reading the same numbers as the rest of us? Because I didn't see a "crash" in EBITDA. In fact, T-Mobile's EBITDA and margin appears to be higher than Sprint's at the end of this quarter. It dipped a bit, true, but it wasn't a significant drop. Sprint's drop was more drastic, as it fell quite a bit. Not only that, T-Mobile will start recognizing the return on ETF buyout (and resale of traded in devices) investments next year. There's also the MetroPCS CDMA network shutdown savings that will improve the cost structure, too.

600MHz banding is up to the FCC. Depending on how many blocks are created, the 3GPP may be forced to create more than one band class because of the varying duplexes. That being said, the FCC already mandated full interoperability across the entire band. If it works out to be 1, 2, 3, or more band classes, they all have to be supported on every 600MHz capable device. As for supporting 3GPP2 by GSM operators, it doesn't make a lot of sense to do that. For one, that automatically locks them down to Qualcomm, which is not something anyone wants (a large part of what makes 3GPP-only devices more cost-effective is the huge number of options for modems and SoCs). While most devices do use Qualcomm in the U.S. now, it's unlikely that it will remain the case going forward. The next generation of GSM/WCDMA/LTE modems and SoCs will be ready in time for the 2015 and 2016 device cycles. We're already starting to see alternative modems being used in devices again (Icera on Nexus 9, Intel on Galaxy Tabs and Padfone, etc.). A final note on this: 3GPP2 has been winding down operations for the last 18 months. Most "members" don't even participate much (if at all) anymore. The CDG dissolved itself last year and re-established itself as the "MDG" in June, focused on migrating 3GPP2 operators to 3GPP technologies. When Verizon Wireless removes CDMA mandate in the summer/fall 2015 requirements, we'll see CDMA device costs shoot up dramatically as Verizon winds down CDMA procurement. This is when we'll start seeing huge problems for the remaining CDMA operators in the US.

T-Mobile's billing system will attempt to provide a discount to match the pricing of the overall account within $1-2 per month if it had to rise, or choose an option that would drop the price radically, and it would attempt to match plan configurations to get as close as possible to a Simple Choice plan. It would then notify you of the upcoming change. Depending on how the system calculates your configuration and situation, you'll be given a Simple Choice Classic or Simple Choice Value plan. Originally, you always got Simple Choice Classic (which is the same as Simple Choice Value, except JUMP and EIP aren't available because you have subsidy instead), though now you may get Simple Choice Value and an even more substantial discount to bring it below your old pricing with the subsidy cost factored out. At this point (with the newer billing system), it may just default to Simple Choice Value now. The problem with eliminating subsidization for Sprint is that there's no major benefit to the consumer to do so. Sprint's network is very tightly locked down (as all CDMA carriers are) and you can only use Sprint devices approved and activated by Sprint anyway. Sprint devices aren't usable with other carriers by design, so it has low resale value. On top of it, the 3G and 4G network performance isn't quite to the level of other carriers (AT&T and T-Mobile's 3G networks offer nearly 30 times the average performance of Sprint's, most carriers offer double the 4G data performance of Sprint, and eCSFB continues to create problems with voice calls). At the end of the day, it's very difficult for Sprint to make itself look appealing.

This is not what happened. People who were on plans other than Simple Choice were converted to Simple Choice plans. There are actually two variants of Simple Choice: Simple Choice Value (no subsidy) and Simple Choice Classic (with subsidy). People who kept their legacy plans are being migrated to Simple Choice Classic automatically as their contracts expire. When they upgrade, they can choose to move to Simple Choice Value. But, they don't have to. This is why T-Mobile doesn't expect penetration of Simple Choice Value to go much beyond 90% (it's at 85% now). Simple Choice Classic will remain around for folks coming from legacy plans who don't want to move up to Simple Choice Value plans.

In M2M, downstream companies massively inflate the MRC (monthly recurring cost) to their customers so that they make a substantial profit. This allows them to take a small loss on hardware for the initial term, if necessary. One firm I know about charges its customers about $15/mo for the "airtime", but its own cost is less than a dollar per month. Depending on the company and how the hardware is sold, these numbers may differ quite a bit. There's also a second reason for the inflation. M2M monthly rates generally cover an extraordinarily small amount of data. For example, a carrier like Sprint would offer $1/mo for 500KB of data, with maybe $0.10 to $0.50 per extra 500KB in overages. Downstream companies "buffer" this cost by charging a larger sum per month (setting it up as a flat rate to their customers) and set aside a bunch of it for "unexpected" airtime charges. Firmware updates are usually the culprit here, as they tend to trigger $3-5 charges in a month.

That's not good. From my experience in M2M (I've worked on M2M products before), the average ARPU is somewhere around $2-5 per month. Sprint absolutely cannot live off of that. The profit margin on those connections is pretty good (>90%), but the raw revenue is too low. From what I know, revenue ratios on types of subscribers work out somewhat similar to this: Postpaid contracted phone+subsidy: 1.0x (baseline) Postpaid phone+EIP: 1.25x Postpaid tablet+EIP: 0.75x Postpaid/prepaid tablet+subsidy: 0.5x Prepaid phone: 0.5x MVNO retail: 0.3x MVNE M2M: 0.2x M2M direct: 0.1x Sprint's subscriber mix has changed rapidly toward those that generate substantially lower revenue. This is why Wall Street has punished the stock. This is going to make the cash burn problem even worse, too.

Meh, I just wanted to explain the term "UMTS" and gently point out that we 3GPP folks love WCDMA and HSPA+ and that bad blood isn't needed.

spray tan

At this point, I believe not. Mexico took back the spectrum last year, and it has been reconfigured for bands 7 and 38. The transition to prepare it for auction should be completed in a couple of years.

Well, strictly speaking, while most people refer to the WCDMA variant when they say "UMTS", the name actually denotes four different systems. The WCDMA version is the one most call UMTS (though DoCoMo originally named it "FOMA", meaning Freedom of Mobile[/Multimedia] Access). It's technically called "UTRA" (Universal Terrestrial Radio Access) or UTRA-FDD. UMTS also encapsulates "E-UTRA" (Evolved UTRA, which we call LTE), UTRA-TDD HCR (TD-WCDMA), and UTRA-TDD LCR (TD-SCDMA). The four air interfaces are incompatible with each other, but they do share common specifications and core network technologies. And A.J., we "GSM zealots" (as you call them) fully recognize Qualcomm's contributions to the development of UMTS. That being said, we still despise CDMA2000 for how Qualcomm handles it compared to how it is forced to handle the UMTS family. Also, it is technically possible to implement CDMA2000/AMPS/D-AMPS with a GSM core. Specifically with CDMA, I think A.J. brought it up once before: GSM1X. The idea is perfectly sound, and it is in use today with China Telecom's CDMA network (China Telecom used to be dual GSM/CDMA before the government forced the reorganization of the mobile network owners). Finally, A.J., there's enough bad blood on both sides. Don't add to it.

Yes. Trefis also justifies it by saying because T-Mobile got those kinds of gains, Sprint will too. They're horrible and unreliable. I've seen similarly stupid analyses from them on other companies... They are also the only firm to make such a statement. Out of more than a dozen analyst firms!

5 MHz Band 2 LTE along I-59 near Picayune, Poplarville, Purvis, just north of Hattiesburg (adjacent to a band 4 site), Laurel, Sandersville, Heidleburg, and Vossburg. On US-45 north of Meridian, Lauderdale, Shuqualak, Macon, and Brooksville are all 5 MHz Band 2 LTE as well. According to Sensorly and RootMetrics it looks like some places like Collins, Mendenhall, Moss Point and Gautier as well. Also worth noting that they deployed some Band 4/UMTS markets in MS within the last 9 months in West Point, Columbus, Oxford, and Vicksburg as well. Also, 5MHz FDD Band 2 LTE is active along Clinton-Raymond Road. Spots of 10MHz FDD Band 4 LTE and WCDMA carriers (two AWS, one PCS) are live within Clinton proper, and it smooths out to ubiquity in Jackson.