Conan Kudo

Dry heaves

Ah yes, I understand now. I didn't realize there were markets where they had more than 10MHz to deploy. I certainly wish more people understood the performance Sprint is capable of in high population areas like cities simply because of the vast Clear spectrum holdings. I honestly cannot wait to see the day when smartphones are using 20GB of data a month and all the other providers literally choke with the limited spectrum for new carriers. It has nothing to do with Rochester yet, but it'll be awesome to see a Sprint that says to its customers "we want you to use our network because it can handle it, unlike those other guys!" Maybe in about five years, if that, at the rate things are going? NV 1.0 to being Sprint alongside its competitors, NV 2.0 to blow them away. On another note, now that the weather has broken back into a little bit of a warm spell, do you think we ought to see some more backhaul upgrades and therefore 4G acceptances around? I am also curious as to whether or not most of the sites that have been done have had fiber backhaul installed versus microwave. Is microwave more of an out of the city type thing? I imagine that would be less common but also less dependant on weather? Sent from my Nexus 7 using Tapatalk To be clear, those 20MHz allocations don't mean the same thing. Because AT&T, Verizon and T-Mobile use FDD allocations, the entire 10/15/20 MHz is used for downlink, since another set of 10/15/20 MHz is used for uplink. This gives them a slight edge over Sprint's TDD 20MHz network in terms of downlink performance, and a slightly more significant edge in the (oft-derided) uplink performance. Under "ordinary" circumstances (where a TDD carrier is set up to be "equivalent" to an FDD carrier), a 20MHz TDD carrier would be equal in performance to a 10MHz FDD carrier (if not slightly lesser due to temporal guards). However, the advantage of TDD systems is that equivalent allocations for downlink and uplink aren't required. In fact, TDD systems are least efficient at 1:1. Depending on the operator's needs, the TDD system can be set up to be downlink dominated, uplink dominated, or a mix of the two depending on the needs at that particular moment. In a downlink dominated scheme, the temporal scheduling factor is set up so that more "time" is allocated for downlink operations than uplink operations. This improves the effective bandwidth of the downlink channel and improves the downlink performance considerably. This allows a 20MHz TDD carrier to offer better downlink performance than a 10MHz FDD carrier. Depending on how extreme the ratio is, you could achieve nearly the downlink performance of a 15MHz FDD carrier. The cost of this scheme is uplink performance, which many believe is an acceptable trade-off. In an uplink dominated scheme, this is completely reversed. Deployments with uplink dominated schemes are rare. One particular use case for this would be for LTE-based IP streaming (where studio cameras are hooked up to LTE and are sending video data over the Internet to another destination). Let's construct an example where this would be useful. Let's say Sprint was contracted to provide service in an area where traditional connection options are unavailable for doing on-location filming. Sprint could set up picocells with high quality backhaul to provide an uplink-dominated TDD network for the purposes of the filming. This would make filming easier when operating in "complex" environments. Sprint has enough 2.6GHz spectrum that this is possible to do without turning down the regular downlink-dominated TDD network. It would just have to control the provisioning so that only the camera devices could connect to it. There's also the ability to change the ratios in a given area based on the need. For example, if an emergency is occurring somewhere that emergency services use the Sprint network, Sprint can temporarily switch the area to uplink dominated to allow dispatch and other half-duplex systems to work more effectively. There is a lot of potential with LTE TDD systems that we haven't quite figured out yet. But one thing is for sure: LTE TDD is really only useful at the high frequency bands, where you can have large spectral allocations. Since Sprint has that in spades, it's definitely good to go on that front!

Crummy Tummy

Red Rum

That's not really "getting" T-Mobile. Besides, network sharing is definitely a good option for both companies that would preserve the competitive pressures of both brands on the incumbents. It also allows both SoftBank and Deutsche Telekom to retain influence in the US market. As for building a network from the ground up, that isn't as hard as it used to be. The US is littered with cell sites owned by tower companies. Dish just has to make deals with these tower companies to get access to the needed site locations for its gear. It would be an expensive venture, but it is doable.

You're discounting small cells. Deploying picocells across the stadium can offer similar (or better) capacity to Wi-Fi APs, and can be co-located with Wi-Fi APs, too. In the case of Sprint Spark, picocells supporting a 20MHz TDD carrier would offer similar capacity to Wi-Fi, perhaps slightly better. And picocells can be sectorized as well (though it is more challenging to do).

It looks like Verizon is including the soft-launched LTE zones that TMUS has scattered throughout the country. For example, the blobs in central Mississippi and the west edge of Texas are not in areas currently announced by T-Mobile as having LTE. But they are indeed there. That being said, T-Mobile does have the smallest LTE footprint, announced or otherwise.

How can I dispute what I've never used? I've only managed to used Band 25 LTE in the last week before returning the Moto X. The low latency (~80ms) was impressive, but the throughput is depressingly low, averaging around 4Mbps, despite the low number of Sprint subscribers in the area. I also had some difficulty maintaining a connection, even in a strong area, but I'm not sure if that was a problem with the Moto X. I am hopeful that if Band 41 service were to launch here, the speeds would be raised to >15Mbps (meeting or exceeding T-Mobile's HSPA+21 average speed). And seriously, why the heck are you mad at me?

I want both Sprint and T-Mobile LTE service in Starkville. Honestly, I don't care too much about LTE from T-Mobile, as the HSPA+ is plenty fast. But I do care about LTE from Sprint, because the EvDO is horrible.

Hardly. I refuse to make any judgement on Spark until I can witness it myself. The only judgement I can make is that I am not likely to be able to try it for at least a year or two, since it'll have to make its way down from Tier 1 cities to Tier 3 cities. I'm hopeful that I'll be able to travel to a Tier 1 city soon to try it. I'm only saying what most of the people I talk to say. While I was forced to cancel my Sprint service again and return my Moto X because of costs (I needed cash to pay for new glasses, since they broke on Christmas), I do plan to re-establish service soon. I do currently still own a Galaxy S4 mini, which is a tri-band device. I would have rather sold it to buy a Nexus 5, but it'll do. And LTE multi-carrier will help, in places where Sprint actually has Band 41 deployed. The number of places it is accessible is growing almost every day, but it's still not going to be very large before the end of the year.

Err, actually, he did mention that. T-Mobile does generally have a 20MHz network in most markets, when you realize he's talking about 10MHz FDD (10+10 MHz). He refers to Verizon's AWS network as a 40MHz network, so that's a good clue to note. He did also mention the lack of mobile broadband coverage outside the metro areas, too. Sprint is mentioning 50-60 Mbps in its press releases. Those get repeated everywhere. It's only after you actually contact Sprint PR for details do they mention the 5-12Mbps number. I don't specifically recall any mentions of 120Mbps outside of Dallas, where it exists today. They've been using Ookla data to show T-Mobile is the fastest, on average.

The problem is that even with all these upgrades, Sprint is perceived to be unable to keep up with the rest of the national operators. That, in itself, is a huge problem. Sprint's Spark isn't helping, it's making everything worse by raising people's expectations and having them crash and burn through real-world experiences.

Category 4 UE won't have carrier aggregation. All LTE CA devices are Category 6 UE. And all UE categories are done with the assumption you are using a single carrier channel. Carrier aggregation throughput differs from the maximum defined throughput for a UE category.

That's LTE multi-carrier technology. Not all vendors have support for it. Maybe all Network Vision vendors do, but I've only confirmed NSN and Samsung gear to have it. I believe the problem is substantially more evident with 1.4MHz/3MHz to 5MHz carriers. If I remember correctly, the overhead subcarrier set size does not change based on the total carrier size. This means that the ratio of data subcarriers to overhead subcarriers becomes a problem at small carrier sizes. And note that I was comparing a wide carriers against aggregated carriers, which it will beat on downlink performance because carrier aggregation necessarily cannot offer the same performance that a wide LTE carrier can. This is because each carrier being aggregated has its own set of overhead subcarriers, which subtracts from the overall available data subcarriers.

That is only true if multi-carrier technology is enabled on the cell site. Otherwise you have to hope that the UE would pick the right carrier to get good performance. As far as I know, only Samsung cell sites have it. No to both. Multi-carrier (load balancing across multiple independent carriers) provides a linear capacity boost while not providing any bandwidth improvements. Carrier aggregation (bonding multiple independent carriers together) provides slightly less than linear boost to downlink performance and capacity. However, wider carriers provide a noticeably higher than linear boost in downlink and uplink performance (compared to aggregating multiple independent carriers covering the same amount of spectrum), and capacity is experiences a higher than linear boost because the ratio of data subcarriers to overhead substantially increases. This is why wider carriers are preferable to multi-carrier and carrier aggregation.

Sprint still tells me that they have no ability to unlock devices it sells.

Pocket Monsters

Orange Box

I don't disagree with that assessment to some extent. However, it is also up to Son-san to make the choices needed to put Sprint on a path to growth. That will require taking a hard look at Sprint's leadership and where Sprint plans to go and reworking them as necessary.