WiWavelength

4.52"

Wow, Ben measured himself to the hundredth of an inch. Now, that is precision.

AJ

Wouldn't the modems be shown on some diagram from the FCC filing?

No, internal processing is not the purview of the FCC. The FCC is concerned primarily with external radiation. So, typical FCC OET filings make no mention of the the actual baseband radio.

AJ

Does the EVO 4G LTE contain an unmentioned second Qualcomm modem so as to enable SVDO? Or does the S4 MSM8960 internally contain multiple radios, such that it can do SVDO by itself? Those are the greater questions right now.

AJ

Forget it! I've been trying to kill them and the stupid things don't want to die.

Note to self: never be houseguest of Josh. Might try to kill you.

AJ

I'm curious: Why make a device compatible with 3G EVDO on 800 SMR when you have no plans to put EVDO on the SMR Band to begin with?

EV-DO 800 capability is already baked into the chipset, power amp module, and antenna configuration. Locking out the capability would require more work than simply letting it go unused.

AJ

Forget that. These ants are marching 20 by 20!

Are you being a good host? The least that you could do is play some music for them.

AJ

Only the older QCELP 13k rates took more than a single 9600 bps data frame to transmit.

Under good signal conditions, 13K QCELP is still the best sounding codec -- in my opinion. It does not utilize a higher sample rate (still 8 kHz), but its higher bit rate does seem to preserve more higher frequency information than does EVRC.

That said, for many years now, few CDMA1X networks allow native subs to use 13K because network operators know that their native handsets are all EVRC capable. Roamers, however, generally get the benefit of the doubt, and the network will honor their specified codec preferences. For example, I have adjusted the codec selection on a VZW handset while roaming on Sprint. With each selection -- 8K QCELP, 13K QCELP, EVRC -- the Sprint network honored the codec preference and set up the call with desired Service Option (SO1, SO32768, SO3).

AJ

Right. Clearwire is not even close to representative of what WiMax is capable of.

The fair comparison would be in equivalent spectrum (both in propagation and in bandwidth). And 10 MHz x 10 MHz of Upper 700 MHz spectrum is certainly not equivalent to 10 MHz (or even 5 MHz) of BRS/EBS 2600 MHz spectrum.

AJ

No. But what would WiMax give at this signal? It's no wonder LTE won out over WiMax.

Scott, come on -- that is not even a remotely relevant comparison. Honestly, you should know better than that.

AJ

Yep. LOL This is at -121

That is probably not RSSI but RSRP (or another LTE specific signal strength measure). If RSRP, one conversion factor that I have encountered is to add 14 dB.

AJ

1xA's default vocoder is EVRC-B, which is the standard for Verizon in most areas. Some people prefer it over EVRC, and it does sound very good, but I personally prefer EVRC. AT&T uses an AMR vocoder though upgraded from GSM's horrible original versions, it's still very, very poor in comparison. There are newer AMR versions with "HD voice" capabilities that Verizon plans to use when rolling out VoLTE this October.

It is important to note, however, that EVRC-NW "HD Voice" does not require Network Vision nor CDMA1X Advanced; EVRC-NW is based on the same 9.6 kbps Rate Set 1 as is EVRC. "HD Voice" comes from greater voice data compression, not from higher rate voice data. So, as long as the BSC/MSC has been updated to decode, transcode, or pass EVRC-NW as necessary, then "HD Voice" should be functional even in those markets that Network Vision has not yet reached.

AJ

I was thinking along the lines for example that you had a 10x10 chunk of spectrum and you allocated it so that 7x3 and then 3x7 with the 7 being the downlink.

I am not quite sure what you are proposing ("7x3 and then 3x7 with the 7 being the downlink"), but the reality of the situation is that a 10 MHz x 10 MHz FDD block is a 10 MHz uplink paired with a 10 MHz downlink. In FDD operation, the uplink and downlink must be separated to avoid horrible adjacent channel interference that could lead to receiver overload.

AJ

Simultaneous 3G and voice would be nice for fallback, but if it diverts resources from more long term efforts, it makes sense for them to skip it. If it could be done at minimal cost as a part of the upgrades, that'd be one thing, but it sounds like that's not the case.

As has been stated many times, SVDO is not a network upgrade, so it does not add cost nor divert resources. Rather, SVDO is a device capability, typically enabled through separate modems for CDMA1X and EV-DO.

AJ

I don't think you understood what I meant. What I was suggesting was if Sprint were to buy out SouthernLinc and whoever else uses that spectrum in the 810-817 MHz so that Sprint could expand its current 800 MHz holdings. It never was meant to be something that Sprint would be able to do now but thank you. I understand that the current spectrum Sprint owns is the ESMR band which is from 817-824 MHz.

Eric, do not take this the wrong way, but I think that you need to let this dream die. That Sprint could buy out and/or relocate SouthernLINC, other ESMR licensees, and (most importantly) public safety users is a pipe dream that has no chance of happening within the next 10 years. Public safety rebanding has been such an ordeal and is still ongoing that we will not see another 800 MHz reconfiguration effort for a long time.

AJ

One thing I noticed that is VERY interesting is that both VZW and AT&T only have 12.5x12.5 in the Cellular band (*except in a few areas) while Sprint has 14x14 in most areas for the SMR.

No, Sprint holds 14 MHz total -- max 7 MHz x 7 MHz ESMR in the reconfigured SMR 800 MHz band. And, in many markets, Sprint is licensed or has available considerably less than 14 MHz.

AJ

"Samsung Galaxy Note shows up on Sprint website"

http://www.mobilebloom.com/samsung-galaxy-note-shows-up-on-sprint-website/2212496/

AJ

I wonder how much carrier nit-picking is involved in all this. I would assume VZN would be upset if AT&T got AWS support but VZN got left out (granted AT&T absolutely needs AWS whereas VZN does not).

Actually, neither "needs" AWS 2100+1700 MHz capability. VZW currently holds AWS across only the eastern half of the US and will wait for the SpectrumCo-Cox AWS transaction to shake out before VZW does anything in AWS. And AT&T, while it has released all of its LTE devices thus far with AWS, quite possibly will never deploy anything in AWS. AT&T's AWS holdings have always been overstated. Now, after assigning much of that AWS to T-Mobile as part of the merger break up fee, AT&T has downright little AWS left in the top markets. See my graph:

AJ

Can LTE be allocated in say a 7x3 configuration? I say that because typically my speeds on evdo are about 50-100k down and 700-800k up. I wouldn't think that all that spectrum would need to be allocated for upload. Most people are used to having under 1mbps for uploads.

A 3 MHz x 7 MHz configuration? No, not really. You may be confusing a 10 MHz FDD license with just a 10 MHz single block of spectrum. In actuality, it is 5 MHz x 5 MHz -- uplink and downlink -- separated by an offset.

Also it looks like LTE is up to 12 times more efficient that EVDO so that should help for the near term capacity problems.

That figure likely comes from comparing LTE 64-QAM (6 bits/symbol) to EV-DO 16-QAM (4 bits/symbol) and LTE 8x8 MIMO to EV-DO SISO on a Hz per Hz basis. Do the math: 6 ÷ 4 = 1.5 and 8 ÷ 1 = 8, then 1.5 × 8 = 12. But that is a highly optimistic comparison; real world performance will not even approach that 12:1 ratio because 64-QAM is usable under only the best radio conditions and 8x8 MIMO will not be available in very many devices ever.

A more realistic comparison is LTE 16-QAM to EV-DO 16-QAM, LTE 2x2 MIMO to EV-DO SISO, and LTE 5 MHz to EV-DO 1.25 MHz. Again, do the math: 4 ÷ 4 = 1 and 2 ÷ 1 = 2 and 5 ÷ 1.25 = 4, then 1 × 2 × 4 = 8. That equates to an 8:1 ratio in favor of LTE, but note that LTE requires fours times the bandwidth to achieve that advantage. So, the real spectral efficiency advance for LTE comes largely from multiple spatial channels, ergo MIMO.

AJ

It would be easier since the new iPad is already at&t & Verizon LTE compatible.

Nope. You cannot overlook the size difference. The iPad is a 9.7 in tablet -- not an apt parallel for the iPhone. For example, that a Toyota Tundra can easily tow a boat does not mean that a Toyota Prius can easily tow a boat.

AJ

The at&t iPhone will be easy as all they have to do is add LTE compatibility, and the same for the Verizon one.

Josh, how will adding LTE capability to the iPhone be "easy"? An LTE iPhone will have to retain all of the GSM/W-CDMA/CDMA1X/EV-DO connectivity of the iPhone 4S plus add 2x2 MIMO LTE 750 antennas for VZW and/or 2x2 MIMO LTE 700/2100+1700 antennas for AT&T. Even if Apple does two LTE models, a la the "new" iPad, it must do so in a phone approximately one seventh the volume of the iPad. Again, how is that "easy"?

As for Sprint iPhone LTE compatibility, I would wager a guess that it will be limited to LTE 1900. And, if the iPhone 4S is any indication, it also will not support CDMA1X 800. And that could be a good thing, as it would confine the throngs of iPhonerds to CDMA1X/EV-DO/LTE 1900, leaving more CDMA1X/LTE 800 and/or TD-LTE 2600 capacity free for other users.

AJ

I wonder why doesn't microwave their own fiber when they can?

Instant oatmeal is a good way to microwave fiber.

AJ

okay. I just assumed that if they were going to go to 4x4 that they would go ahead and do so now on the towers...that is as long as its downward/backward compatible...Same thought for LTE-Advanced, I was assuming that the necc hardware was there on towers after NV upgrades take place and it would be something that could be flipped on via the tower easy once handsets are out to support the feature later...

Sgt., I do not keep up with standards ratification and equipment availability as much as I should. But, as far as I know, LTE Advanced is not yet ready for prime time. So, LTE Advanced was not an option -- at least, not at the start of the Network Vision sweep, much like LTE was not an option at the start of the WiMAX deployment.

On my soapbox, I think that infrastructure vendors stagger their R&D with planned obsolescence in mind. Why sell wireless carriers LTE Advanced directly when they can sell them LTE first, then sell them LTE Advanced a year or two later? Planned obsolescence is one of the most despicable aspects of tech industry capitalism.

So all in all they are doing the one NV rollout upgrades and then in a sense for LTE-Advanced they will have to do another nationwide set of tower upgrades in the near future?...

I believe that is true, but only to a certain extent. So, do not get carried away. For example, Network Vision is a complete overhaul of antennas, base stations, and backhaul. An LTE Advanced sweep, on the other hand, will be a modest evolutionary update, not a revolutionary change. Because, other than carrier aggregation, the biggest changes from LTE to LTE Advanced are on the uplink, so that falls more so on the device manufacturers.

AJ

Oh okay so when they said "deploy" they meant as in, installed in handsets, not deployed on towers...I took it as it was something they had to change on the towers.

"Deploy" could mean some updates to base station infrastructure. I do not know, as I am not certain that current Sprint LTE hardware supports greater than 2x2 downlink MIMO. Even if an infrastructure upgrade is necessary, there would be little reason to install the hardware for 4x4 downlink MIMO now because only LTE UE category 5 (see table below) devices can take advantage of 4x4 downlink MIMO. But, to my knowledge, all LTE devices thus far have been category 2 (Motorola) or category 3 (most/all other OEMs). So, category 5 devices may not hit the market anytime soon. Furthermore, Sprint's roll out of 4x4 downlink MIMO appears to coincide with its roll out of LTE Advanced; hence, an infrastructure upgrade will be required either way.

AJ

So I have a question as well. I assume the Evo LTE will use 1x Advanced for HD Voice. If they are already advertising this now, does that mean they will be deploying 1x Advanced during the first round but maybe on a 1900mhz carrier (a through f)

No, CDMA1X Advanced is not required for HD Voice, which is technically the EVRC-NW codec. EVRC-NW has greater frequency response, but stronger data compression techniques keep its max voice data rate at the same 8.55 kbps as the current EVRC codec.

AJ

Lastly in response to that article above which stated they were currently deploying 2x2 MIMO and later going to 4x4 MIMO....reasoning why not going straight to 4x4 MIMO with the NV rollout upgrades?

The simple answer is that 4x4 MIMO requires four downlink antennas. Some larger handsets may eventually be able to accommodate 4x4 MIMO, especially for higher frequency bands. But, more likely, 4x4 and 8x8 MIMO will be largely limited to tablets and laptops, which have much better form factor for multiple, spaced antennas.

AJ