boomerbubba

That depends on which map you are looking at. There are some maps for each market showing all the towers that will be converted during the life of the project, which is essentially every Sprint tower. There is a separate map, updated periodically, showing "completed" towers. Not all completed towers everywhere are live to customers, but those in Atlanta presumably are because that market has been launched.

I wonder what happens on AT&T LTE phones. Since the Android API provides no LTE-specific information beyond signal strength metrics, and no LTE location or IDs at all, whatever the OEMs do under the hood is sort of a mystery. Does HTC document this "LTE Record" screen at all? I wonder if HTC is just displaying the only data that Android happens to make available from the CDMA location API. There also is a GSM location API that provides the various cell IDs, etc., defined for GSM. But there is no LTE location API defined for Android yet. EDIT: One thing that make me suspicious that the HTC ECO is not doing anyreally deep under-the-covers OEM magic with its "LTE Record" screen is this: All that SID/NID/BSID data is also reported by Netmonitor, which is just an unrooted app using the vanilla Android APIs. And that data just comes from the CDMA location API. I have looked for some LTE documentation that would explain in the abstract what system of IDs it even uses. Gotta believe there is some unique ID somewhere. Can you point so some document that explains that, or how hybrid networks ID their sites?

I thought that SID, NID and BSID are CDMA IDs, not IDs defined for LTE. (I even thought I read you saying something like that.) Are these IDs found in the LTE standard? Would an LTE-capable GSM device capture these IDs, too?

Those are CDMA base station IDs. They may or may not be colocated on towers with functioning LTE.

Spint just announced an OTA today for your GS3. Release notes say that among other things, version L710VPLG4 includes:

This was posted in another thread, but it deserves notice here. Sprint announced an OTA update today, version L710VPLG4, which among other things promises:

The item in the release notes that most GS3 users in LTE markets are probably most interested in is this one: Lets hope it cures the LTE/EVDO threshold problem on this model.

You might just start reading at the top of this thread. That article as linked in the third comment.

FYI, I bumped into an online Sprint "announcement" today on LTE rollout progress. It did correctly (and belatedly?) include a caveat: Also some links to selected trade press coverage.

Actually, I already looked at that area (San Bernardino). See my comment above. Open Signal's mapped towers there don't correlate to actual tower sites well at all.. I think this is a good example of experimental placebo effects. Users invest time using this app, so they want to believe in it no matter what. (And most end users don't have our advantage at S4GRU -- master Sprint maps of the tower sites.) The field crowdsourcing method is an interesting idea for plotting empirical signal-strength coverage in the corridors where the handsets themselves happen to travel (although I like Sensorly's site better for that purpose in the maps I have browsed.) But Open Signal's methods attempting to extrapolate that same data to impute tower locations just don't work very well.

Okay, I guess iHeads are welcome to play, too. Is there some website we all can look at describing these iOS apps?

I'm bending the rules to look this up for you, but other readers deserve to know how off your description is. According to the authoritative project maps, there is a Sprint tower at 30th & Mountain View, but not on the Open Signal Maps website. (Actually, as I compare that website with the master S4GRU map of the area, I find no decent correlation between the various "towers" mapped by Open Signal and the actual locations.) And there is no Sprint tower at Electric Ave. & Pershing. I can't explain what Sprint customer tells you, but I wouldn't be surprised if they agreed with your geographic analysis just to get you off the phone. That is not uncommon behavior among frontline CS reps, in my experience. This strikes me as a rather impertinent question from a newbie guest, but then this is not my site. I'm a guest here myself. But like other sponsors, I choose to chip in some to help pay the costs. FWIW, you can read about sponsorship here.

The only tower sites I can find mapped on Sprint's website are a handful of "Recent" or "Future" 3G towers. These are a tiny fraction of Sprint's main-network antenna sites, of which there are about 40,000 (not including IDEN sites that will be decommissioned or Wimax sites, which are in Clearwires's own separate network.) I recall seeing some other maps on Sprint's site showing some towers scheduled for maintenance upgrades, but they are also relatively few. The only place I have ever seen complete and accurate maps of Sprint towers are in S4GRU's sponsor forums.

Could you post some screenshots of that? I haven't seen such a thing here or in other forums, but I have seen several examples of wildly inaccurate "tower" plots by Open Signal Maps. The developers even warn about errors on their own Google Play page: So there may be such a unicorn as an accurate Open Signal Maps plot of Sprint towers. I just haven't seen one. And if you do post a screenshot, could you explain how you confirmed that these "tower" sites were correct? Using apps that track the broadcast CDMA coordinates, I have always been able to discern which tower I was connected to and where it is, even if it takes some work in the case of the many towers near me that squawk offset coordinates. But for this exercise I also have an asset you probably don't have, since I notice that you are not an S4GRU sponsor. I am, so I have the Rosetta Stone -- Robert's authoritative maps of where the Sprint towers actually are.

One problem with these articles is that they conflated technical issues with business issues. I don't give much credence to their not-enough-spectrum theory, and I think the technical merits of the NV project's state down the road are attractive. But I have to think Sprint's overall financial and business outlook is an open question. It's all about timing, and Sprint has to try holding on to subscribers while the NV buildout proceeds. I am neither a Sprint fan nor a hater, just a rational customer. And I have to say I would be hard pressed to stay with Sprint if I did not expect LTE coverage coming to my area relatively soon. (I do expect LTE pretty soon, thanks to this site. But many others are not so fortunate.) So while I join in hoping for Sprint's success because it is in our mutual interest, I recognize that success is not guaranteed. The carrier might lose revenue faster than it can rebuild its network.

Become a sponsor.

Try CDMA Field Test to find CDMA connections. Just be aware of the anomaly about some towers squawking three different coordinates, one for each sector BSID. If the coordinates mapped by the test app don't match a tower site on the S4GRU map, then this apparently is the case for yours. You might have to explore around the towers to find the three sectors with sequentlally numbered BSIDs.

Interesting. Robert's hypothesis (explained here) is that the threshold settings on the handsets are affecting how they downshift from LTE as a function of signal strength, and not benignly. If that hypothesis is valid, then different OEMs' handsets may very well calculate the threshold and downshift differently.

Understand first that there is no app known around here that will map LTE towers directly. Some apps purport to do so, but they are misleading. There are some apps that will map CDMA sectors, each with its own BSID (Base Station ID) from the coordinates broadcast by the tower transmitters. Sometimes those coordinates are the actual tower sites. But sometimes [sponsor-only link] each sector on a given tower squawks coordinates that are offset some distance away (maybe even miles). We don't know why that is. But since you as a S4GRU sponsor have available to you maps of actual Sprint towers, you can usually tell be trial and error which case is which: Either the coordinates are exactly at a tower site, or they are grouped in a triangle pattern around it, with BSID values that are sequential. As for which app to use for that purpose, I don't like Netmonitor because I don't trust its permissions for privacy reasons. I prefer CDMA Field Test, which will only map a single BSID at a time. But is does support capturing all the IDs and coordinates in a log, which is exported as a KML map file or as a flat file you can use to build maps yourself. One of the worst apps for "tower" mapping is Open Signal Maps, which really just computes a guess about tower locations by trying to triangulate crowd-sourced signal-strength data. This method can be wildly inaccurate. I have never seen it pinpoint a real tower.

They launch hidden utility apps. Such apps often can also be invoked by making a shortcut within a third-party launcher such a LauncherPro in lieu of Touchwiz. Yet another method that might work is a standalone laucher app such as QuickShortcutMaker, which seems to work with many hidden utilities on my Samsung Epic 4G. (I don't have a GS3.) These methods might launch some LTE utilities on the GS3, which we don't know much about. It may be that this is a route to try tweaking the LTE/EVDO fallback thresholds, as Robert has suggested.

This is all news to me, and I have never received an alert. Who is sending these out -- the carrier? Does someone opt in to the service?

Thanks for that addtional info. I guess we can chalk this case up to you capturing a test tower in the wild temporarily. As for as the app screenshots go, neither Open Signal Maps nor Netmonitor actually maps LTE sites, although they might give that erroneous impression. (Heck, Open Signal Maps can't even map CDMA sites at all. It just guesses from crowdsourced submissions.) Netmonitor is mapping your CDMA connection, and plotting the coordinates squawked by the CDMA radio on the tower. In this case, that much is dead-on accurate. That site is a Sprint tower site. But it may or may not be the same tower where your LTE connection was coming from. (Although in your case I suspect that it was, after studying the Sponsor-accessible maps.) But there is no known app capable of mapping an actual LTE connection. You can see LTE signal strength under the Android's general Setttings -> About ... menus. But that is all the device knows. Neither the Android API nor the apps calling it have access to any LTE site indentifiers or locations.

Interesting. This is consistent with Robert's hypothesis that handsets may be downshifting too soon when the LTE signal is not very strong. I notice that yours is a Galaxy S3. First, a caveat: That Netmonitor screenshot plots the CDMA tower, which may or may not be the tower you were connected to for LTE. (We know of no Android app that maps actual LTE connections.) But it is plausible that this is also the LTE tower, which I believe would be the closest to your actual location. As far as I can tell from Google Earth, you were about .66 miles aways from that tower with an unobstructed line of sight. The LTE signal strength was not captured. (That can only be done from the Android Settings -> About ... menus.) But just from the SpeedTest screen showing 12.14Mbps down, I guess that is plausible performance in such an environment. What do the RF experts think? I wonder if the LTE tower station was really turned off, or if this was an artifact of the threshold issue on the phone. There may be two threshold settings -- once controlling downshifts from LTE and one controlling upshifts. Anecdotally, it seems fairly common in launched areas that some users lose their LTE connection, then don't reacquire it. I wonder if the phone would get the LTE back if it was reset by toggling the LTE in the Settings menu, or maybe a reboot. (Of course, LTE on the tower may have been turned off, since the Chicago market isn't launched yet. It may even have been turned off coincidentally as you entered the building. That could explain this whole scenario, but we can't know for sure.)

+1 on both apps -1 on both apps: -1 on Open Signal Maps because its "tower" locations are so inaccurate. -1 on Netmonitor for its possible risk to privacy.

I am experiencing some belated cognitive dissonance when I recall WiWavelength's initial test of the LTE rollout. He seemed to detect a threshold that kept LTE alive with a moderately weak signal -- a condition he deliberately tested. I think he was using an EVO LTE. Maybe the threshold setting vary across different devices?