milan03

If they were equal to full builds Att and Verizon would've done them long ago but the fact that they do not even for extremely rural sites should tell you something about those types of setups.

Didn't Verizon up until last year have LTE700 setup without RRUs? Even now a lot of their LTE700 sites don't have RRUs.

I don't see how low band is going to increase pops by that much. It should bring the LTE network up to GSM 1900 coverage levels or a little more, but it's not gonna increase pop coverage. They'll still have the same amount of markets covered after the 700 rollout.

Considering T-Mobile's decades old and neglected Nortel 1900MHz GSM infrastructure vs a 700MHz spectrum band that propagates at least 2.5x further, which will also require brand new antennas and RRUs at the tower top, I don't think that extra 4 million pops estimation is unreasonable. 

 

Early user reports are already showing improved GSM coverage and 1900MHz LTE that fully matches old Nortel 2G footprint after the simple act of replacing the rusty and outdated cabinets and equipment with new NSN/Ericsson base stations on the ground.

I have a hard time believing their existing 2G-to-LTE footprint covers 290mil pops. Sprint's entire network is only around 270mil pops, if I remember correctly.

 

AT&T's LTE covers 300m POPs. It would stand to reason that T-Mobile's network needs to be roughly the same size as AT&T's to match that number.

 

 

Legere is the one who claimed he would "match Verizon almost everywhere, and win". It's not like I'm expecting him to cover every rural area -- these are places that AT&T and Sprint already have coverage too.

 

He's intentionally setting expectations high, by claiming to match Verizon "almost everywhere", and by claiming to have an LTE network equal in size (by POPs covered) to AT&T's. If he's not actually planning to match most of Verizons / all of AT&T's coverage, he should stop publicly promising to do that.

According to official T-Mobile info, right now the entire footprint is at 286 million pops. With the addition of 700MHz LTE overlay, it's reasonable to guess that reach could be extended by close to 4 million pops.

 

Now obviously, I have no way of verifying how many pops they cover and if they're honest or not, but since they have been publicly stating that 286M number for a while now I'm guessing they're fully ready to stand behind their claim.

 

Also, as I said in another thread, they've already met and exceeded quite a few goals that at first may have looked like a complete BS. If their execs and engineers didn't believe that they can meet 300 million pops goal by the end of '15, I'm sure they would've set the expectations lower so they can meet and exceed. I'm an optimist, but I guess we'll see.

If T-Mobile needs to "fudge" their rollout, I imagine they would do it by leasing a single site in places like Cadillac, but claiming the entire town's population in their "300 million POPs covered" LTE count. (Which would technically be true, but only barely -- and not anywhere near the same quality/usability as Verizon's coverage there is)

I think you certainly have a valid concern as it's not gonna be easy for them to fully match Verizon/AT&T rural cell count maybe ever, but at the same token overlaying the existing 2G footprint with PCS and 700MHz LTE should bring them pretty close to 290 million pop mark.

 

Population of Cadillac, MI is just over 10,000, and it's important to understand that T-Mobile never claimed that they'll actually match Verizon's rural areas cell for cell. Their goal is 300 million by the end of '15, Verizon is already at 303 million pops, and while some rural areas may be left out others will emerge, and we should probably have our expectations reset.

Overall, I'm of the mind that T-Mobile should just leave it up to their ground level engineers in a certain area whether a site gets a ground mount LTE set up like the rural 1900 sites in my area or the full modernization treatment. For most areas ground mount should suffice. Now if there are areas with significant congestion or coverage issues, I'd leave fully modernized sites as an option. The 1900 in my neck of the woods (Perryville, MO) is 10x10 and is faster than AT&T and similarly fast to Verizon. That should give them good headroom for a while.

 

Sprint should consider taking the non-Sparked rural Sprint areas that have low CDMA traffic and most of it on 800 MHz and deploying a 10x10 LTE channel 2 setup along with 5x5 in 25 and 26. That should give Sprint similar headroom to T-Mobile.

Good to hear that they finally have LTE live in your area, and even better 10MHz FDD!

How is T-Mobile going to cover 300 million people with 700 mhz, when it's not even available in 30+ states, and in the states it is there are restrictions or its not statewide?

 

I can't understand how they expect to reach those numbers.

They never claimed that they will cover 300 million pops with 700Mhz LTE alone.

Samsung Electronics Launches the World’s First LTE Advanced Tri-Band Carrier Aggregation Smartphone

 

http://global.samsungtomorrow.com/samsung-electronics-launches-the-worlds-first-lte-advanced-tri-band-carrier-aggregation-smartphone/

 

They really wanted to be the first

The ePDG is part of a larger upgrade that requires migrating Wi-Fi calling to the 3GPP network core and enabling core access over the IP network. Sprint's current Wi-Fi Calling platform doesn't use the 3GPP network core to begin with, so that's already a pretty huge hurdle to overcome.

Yup, that's what had me completely puzzled while reading Euteneuer's transcript, but since he's a Sprint executive I'm going to assume that he knows what he's talking about.

Hate to burst your bubble but don't count on 8.2 to include wifi calling. I believe it will be mostly health kit and Apple Watch related updates.

 

 

Sent from my iPhone 6 using Tapatalk

Yeah that's what it appears to be, but Carrier Updates can be pushed OTA at any time.

 

In order for next gen VoWiFi on iPhone 6 to be supported, the operator is required to upgrade their core network with ePDG (evolved Packet Data Gateway) which itself is a fairly new technology. The most notable benefit is seamless handoff to and from VoLTE.

 

Euteneuer's statement that Sprint WiFi Calling support is coming to iPhone 6 indicates that Sprint is working on these upgrades.

Google Wallet's doing better than Apple Pay, even though carriers like T-Mobile push Isis/Softcard on most of their android devices. 

 

http://www.zdnet.com/article/apple-pay-responsible-for-1-percent-of-november-digital-payments-google-wallet-garners-4-percent/

Don't forget the amount of Google Wallet capable devices vs the amount of iPhone 6/6+ in the US. Google Wallet's been around for a 3+ years.

For what it's worth, I've unlocked iPhones via 3rd parties before and the unlock carried when I had to get it replaced. I just don't know if that's the case for Sprint since they don't unlock for domestic use.

If you've already verified by SIM swapping that your existing Unlock Policy allows you to attach to domestic operators other than Sprint, it confirms that desired Full Unlock Sprint policy actually exists on iTunes servers, and since it's now associated with your iPhone it will carry over to future replacement devices when replaced at the Genius Bar.

Now that it's fully legal again, there are services to unlock Sprint iPhones. I used one that is legit, in that I restored my iPhone and it said "congrats, your iPhone is now factory unlocked." I tried a T-Mobile SIM and it worked perfectly. My question is that should I need to get the phone replaced under warranty and go to the Apple store, will they give me a new unlocked phone for domestic use? Like I said, the phone seems to be completely factory unlocked... Just not through Sprint. The service said it was done through "Apple factory unlock." not Sprint.

 

For those of you that say there are no unlocked domestic Sprint phones that are sold, there are, just through Apple directly if you pay full price.

Once factory unlocked via iTunes, replacement phone that you will get at the Genius Bar will always carry that same Unlock Policy. It's iTunes servers that store that information, not Sprint, so once flagged as unlocked, the future activation chain remains unlocked.

It's right in the

headline: — New High-Performance 64-bit Snapdragon Processors Integrate Cat 6 LTE Advanced / Carrier Aggregation up to 60 MHz —

 

So fine, Qualcomm made a typo.

If anything, that original 300Mbps 3x CA capable integrated baseband from April had lower clocked DSP, while the current one has increased DSP clock capable of more processing bandwidth. It's the same exact 20nm manufacturing process, which means that upgrading the existing baseband silicon wasn't necessarily a difficult thing to do.

 

And finally, there must be a good reason why they've never commercially launched that Cat 6 with 3x CA with Snapdragon. This is obviously much better for the consumer and operator, having integrated MDM9x45 Cat 9 capable IP Stack built into the flagship SoC for 2015. 

And I have posted my own links to show that the 9x35 modem, as implemented in the SD 808/810, supported 3x20 (that's 20+20+20 aggregate capacity, not just 20+10+10), which is a step up from how 9x35 had been added to earlier LTE-A phones with the SD 805.

 

So either the initial PR in April from Qualcomm re the SD 810 had a typo (which then led to errors in all the articles like the Anandtech one that relied on it), or there is something about this we're not getting.

 

In any case, it's largely a moot point and in the past. Going forward, whether the modem in the SD 810 is still some sort of modified 9x35, or the brand new 9x45, I hope that by the time it is released that they will have found a way to enable Cat 10 uplink CA, if they haven't already. Personally, I'd rather be able to utilize two uplink carriers together than a third for downlink.

Yeah, no those links aren't showing 3x20MHz in SD810. There hasn't been a single test conducted with 3x 20MHz using snapdragon 810 prior to upgrade to Cat 9. All those old tests are clearly 20+10+10=40MHz, and for a reason. Cat 6 can only do 40MHz of aggregate spectrum. duh...

 

Ulm, Germany - Nokia Networks and Qualcomm Technologies, Inc., a wholly-owned subsidiary of Qualcomm Incorporated (NASDAQ: QCOM), have completed interoperability testing of LTE-Advanced three-band Carrier Aggregation (3x CA), allowing more operators to achieve subscriber data rates of up to 300 Mbps in their network environments.

 

The tests utilized the Nokia Flexi Multiradio 10 Base Station and Qualcomm® Snapdragon™ 810 processor and allows operators, such as SK Telecom, to offer data rates of up to 300 Mbps to their subscribers using a network of three frequencies (20 MHz + 10 MHz + 10 MHz). 3x CA provides an alternative option to operators to reach 40Mhz carrier aggregation, as this option was previously available only to the operators with at least two bands of 20 MHz bandwidth each."

 

 

No doubt in my mind that original 810 was Cat 6 capable of 40MHz aggregate downlink capacity.

Also, going forward, only if you REALLY have no idea how Qualcomm launches their next gen modems, you'd guess that SD810 is MDM9x35. And I've been telling you why that's not the case , but I guess that wasn't enough, so you continue to summarize your posts assuming 9x35 in there, magically with Cat 9 capabilities. 

 

I'll just drop it right here. Enough is enough.

I don't know man, ask Qualcomm. They were the ones that claimed it could do 3x20 CA as far back as April. That was the question I posed earlier in this thread. My theory was that it was done simply in an attempt to provide product differentiation. Another possible reason could have been that when they first began sampling the chip, tests run at Cat 9 speeds led to thermal issues that only cleared up with time after the node matured, so they initially only felt comfortable promoting it as Cat 6/300 Mbps.

 

 

The SD 808 & 810 supported Cat 7 (w/ 2x20 MHz uplink CA) before, as mentioned in the Anandtech article & detailed in my post from November. If the new version of the SD 810 does indeed include the 9x45, a Cat 10 modem, I'm not sure why the SoC would be limited to Cat 9. If the SD 810 now supports only 50 Mbps uplink, then they've taken an unfortunate step backwards.

 

I've posted a few charts that clearly explain the strategy behind 3x CA with 40MHz aggregate capacity. There is absolutely nothing unclear to me, it's all common sense. Qualcomm's only Cat 9/10 solution is MDM9x45, so that's in Snapdragon 810. Standalone MDM9x45 is the only Cat 10 Solution by Qualcomm at this point according to their loud and clear presser.

 

So no source then. I've already read the pertinent press release, as I mentioned, and it doesn't explicitly say that it is using the 9x45. Other sources that do happen to mention the modem used (dated after the Dec 11 announcement from Qualcomm) still say it is the 9x35.

I'm not saying you're wrong. Indeed, I hope you're right. It'll be cool for the SD 810 to have all the new benefits of the MDM9x45. I would just be surprised if they've managed to do that, since the feature set of chips is usually already locked-in this close to volume production.

Given that the 9x35-based solution used in the chip could already address 60 MHz of aggregated spectrum, it is within the realm of possibility that there was some sort of software-based patch applied to that modem to allow it to support 300 RBs @ Cat 9/10 speeds, so that the SoC's basic components have not changed. The chip will be out in a couple of months anyway so hopefully we'll get an updated in-depth analysis of the architecture soon.

Yeah, well this is common sense. There is a valid reason why a standalone/integrated modem is marketed as Cat 6 and not Cat 9. It isn't Cat 9 because it can't aggregate 60Mhz of downlink spectrum. Cat 6 is limited to 40Mhz.

 

 

 

 

 

No disrespect, but why would you ever assume that anyone in their right mind would go ahead and build a SoC with the integrated modem capable of 450Mbps, but limiting the capabilities to 300Mbps?

 

Aside from that, the benefit of a standalone MDM9x45 solutuon (vs integrated) is CA on the uplink (100Mbps) which makes it a Cat 10 solution. The downlink is the same 450Mbps. SD810 is only Cat 9, without CA on the uplink.

Do you have a source that says that they've switched out the 9x35 modem for the 9x45? The official press release doesn't say either way. A few third-party sites of questionable reliability that I've come across say that the 9x35 is still what is being used.

MDM9x35 = Qualcomm's 4th Gen modem, Category 6 capable of 300Mbps using either 2x or 3x CA.

MDM9x45 = Qualcomm's 5th Gen modem, Cat 9 or 10, capable of 450Mbps on the downlink with 3x CA.

 

Browse through their press releases https://www.qualcomm.com/news/releases

The SD 810 uses a bolt-on solution to extend that to 60 MHz, and was initially advertised as such. That much is clear. It could connect to all 60 MHz, even if it may have only been able to "process" 40 MHz worth of RB's.

What isn't clear, and what we can only speculate about, is why they didn't just run with that all the way & have it certified as Cat 9-ready back in April. Raising the RB limit to 300 to keep up with the competition, rather than being able to keep it artificially lower in order to differentiate it enough from its eventual successor, seems the most likely answer.

Because that's not what it was. It was a MDM9x35 IP stack integrated into SoC. Now they're integrating MDM9x45 that was suppose to be released only as a standalone. Not the same solution.

Yes, the press release did say it was capable of 3x20 MHz CA. The Anandtech article focuses on 20+10+10 only because the author forgot or was not aware of Sprint's plans, and thought that was the widest combination that would available for the foreseeable future within the US.

Let me put it differently. Any LTE Cat 6 modem is capable of up to 300Mbps, and capable of addressing up to 40Mhz of aggregate downlink capacity, period. Thats what Cat 6 is by its definition.

 

So if you are a wireless operator that for some reason have three 20MHz disparate LTE channels and want to aggregate them, that Cat 6 MDM9x35 UE will able to address up to 200 RBs, 40Mhz of downlink capacity, 300Mbps peak downlink rates. Plain and simple.

No. The SD 808 & 810 were always capable of 3x20 MHz CA, as detailed in this Anandtech article, which I referenced in part for this post. Despite that, they had initially only been rated to 300 Mbps (Cat 6/7). I seriously doubt that this late in the process that they have replaced the integrated 9x35 modem with the 9x45. Rather, in the face of announcements from Samsung of upcoming Cat 9 support, and/or as a result of additional testing during the validation process that went well, they decided to up the advertised speed limit to 450 Mbps. With respect to Sprint's network plans, this increase doesn't really make a difference, but it is a rather curious development, and begs the question of what exactly they changed (just the modem's firmware?), and why they hadn't felt compelled to advertise it as Cat 9/10 from the get-go (lack of operator preparedness?)

The thing is that Anandtech article also states that SD810 is Cat 6/7 capable of 20+10+10. It was never capable of aggregating three 20Mhz component carriers, per Qualcomm's pressers. It was the very first 3x CA capable solution, but that's about it. 300Mbps max capable of addressing up to 40MHz aggregate downlink capacity.

 

"Both SoCs have a MDM9x35 derived LTE Category 6/7 modem. The SoCs feature essentially the same modem core as a 9x35 discrete modem, but with one exception: Qualcomm enabled support for 3 carrier aggregation LTE (up from 2). The discrete 9x35 modem implementation can aggregate up to two 20MHz LTE carriers in order to reach Cat 6 LTE's 300Mbps peak download rate. The 808/810, on the other hand, can combine up to three 20MHz LTE carriers (although you'll likely see 3x CA used with narrower channels, e.g. 20MHz + 5MHz + 5MHz or 20MHz + 10MHz + 10MHz)."

Another trial this time in the UK (EE) with Snapdragon 810, in a 20+20+15MHz combination. 410Mbps peak rates.

 

http://www.telecoms.com/312571/huawei-ee-qualcomm-claim-410mbps-via-lte-cat-9/?utm_source=rss&utm_medium=rss&utm_campaign=huawei-ee-qualcomm-claim-410mbps-via-lte-cat-9

I know that the speedtest app creates a stupid pissing contest amongst wireless customers, but I honestly feel that even at the point of multiple band 41 carriers/CA, no user should be allowed max bandwidth. Why not just cap it at a reasonable speed. Let everyone have a fair share of the available bandwidth, instead of one guy pulling 100mbps+ so that he can illegally feed his home ISP needs.

 

 

Sent from my iPhone 6+

That would defeat the purpose and reduce spectral efficiency of the LTE network.

 

In order to keep spectral efficiency high, scheduler dynamically allocates resource blocks based on device feedback, activity, etc. For instance, if a device with low signal quality and low SNR requests light HTML content in an indoor environment unable to ever properly utilize MIMO, but gets equal artificial provisioning (speed cap) as the device that's outdoor with high signal quality and SNR which can easily multiplex and peak higher, that network is effectively wasting their resource blocks, and becoming highly inefficient.

 

The ultimate goal for wireless operators is to move more and more data, increase (bit/s)/Hz or spectral efficiency, and ultimately collect more revenue. That "data hogs" argument makes me laugh every single time, as a properly deployed and managed network is suppose to be able to manage their deployed spectrum resources and/or increase density of their infrastructure accordingly. That's why ToS exists.

 

This is also why at this point in time forward thinking wireless operators all over the world look to densify their networks from the inside out, by deploying micros, femtos, or VoWiFi to complement their existing macro grid, and increase the efficiency.

How is the wideband LTE coverage in NYC?

Coverage is fantastic mainly because of their dense deployment in NYC. Indoor penetration could be better, but VoWIFi comes in handy.

Those are going to be some wonky EARFCNs and UARFCNs.  I would like to see some engineering screenshots.

 

AJ

EARFCN 2236. UARFCN hasn't changed at all.

 

The SD808 & 810 were always going to support the aggregation of three 20 MHz carriers, which if they're FDD would theoretically reach 450 Mbps. So what exactly did Qualcomm change to now label them as Cat 9/10 rather than 6/7? Is this just a marketing change prompted by Samsung's announcement of Cat 9 support in their next Exynos SoC? Why wouldn't they have just announced them as Cat 9/10 to begin with?

 

810 was originally announced with integrated Cat 6 capabilities, and with maximum aggregate downlink capacity at 300Mbps. So you could aggregate three component carriers, but 20+10+10MHz or similar. Perfect for Verizon if the ever decide to do 3xCA.

 

Cat 9/10 is capable of 450Mbps on the downlink by aggregating three 20MHz CCs. That was going to be a standalone MDM9x45 baseband processor, but they've decided to integrate it into SD810 as well.