SignalCheck - Android app to monitor your Wi-Fi/2G/3G/4G LTE/5G-NR signal strengths

[jthawks]

56 minutes ago, Flompholph said:

Triband.

Triband antenas have been found where sprint puts what we guess could be their 4th carrier GCI ending. 

 

Didn't though they were doing that here.  So that's good

[jthawks]

5 minutes ago, jthawks said:

Didn't though they were doing that here.  So that's good

So should I investigate this? 

[Dkoellerwx]

51 minutes ago, jthawks said:

Didn't though they were doing that here.  So that's good

 

45 minutes ago, jthawks said:

So should I investigate this? 

This is standard second carrier for Samsung tri-band sites. As Denver is a Samsung market, it is to be expected.

[mikejeep]

On 4/9/2018 at 4:25 PM, Dkoellerwx said:

This is standard second carrier for Samsung tri-band sites. As Denver is a Samsung market, it is to be expected.

I don't know if it's definitely the second carrier (that should be 03/04/05), but it is an additional carrier that varies by market.

-Mike

[Dkoellerwx]

6 minutes ago, mikejeep said:

I don't know if it's definitely the second carrier (that should be 03/04/05), but it is an additional carrier that varies by market.

-Mike

Based on the EARFCN and it's from the Denver area, it is the second carrier. As far as I've seen, that is the typical ending for a second carrier tri-band antenna site in a Samsung market?

[mikejeep]

I have received a lot of messages over the past 3 weeks regarding improper LTE band identification for T-Mobile users. There is an Android bug that causes EARFCNs assigned to newer LTE bands (>46) to be reported incorrectly, such as Band 66 and Band 71. SignalCheck is showing Band 2/3/5/6/7/etc. because that is what the OS is reporting.

For more details, please see this bug report; I encourage everyone to 'star' it and/or adding your own report to try getting it more attention: https://issuetracker.google.com/issues/37136986

I am trying to implement a workaround that will at least help T-Mobile users until Android fixes the issue.

-Mike

 

[mikejeep]

2 minutes ago, Dkoellerwx said:

Based on the EARFCN and it's from the Denver area, it is the second carrier. As far as I've seen, that is the typical ending for a second carrier tri-band antenna site in a Samsung market?

I have Samsung B41 second carriers as 03/04/05, but perhaps things changed? The "TB" indicator was implemented (and has been discussed a few times) because it wasn't consistently deployed as a specific carrier.

-Mike

[Dkoellerwx]

2 minutes ago, mikejeep said:

I have Samsung B41 second carriers as 03/04/05, but perhaps things changed? The "TB" indicator was implemented (and has been discussed a few times) because it wasn't consistently deployed as a specific carrier.

-Mike

Yes, on a normal 8T8R site, 03/04/05 is the normal second carrier. On 4T4R tri-band antennas in Samsung markets, 09/0A/0B has been used fairly regularly. I do not know off the top of my head what is done with the 3rd carrier on the tri-band antenna sites (normal 8T8R being 06/07/08, which is already covered).

[Trip]

Mike,

1) As indicated previously, in my area (DC) which I believe is a Samsung Clear market, 03/04/05 are third carrier, but SCP currently doesn't note it as anything other than standard B41.

2) Yesterday, I finally managed to send you diagnostics showing "LTE" without a band indicator while connected to a Clear site.  I sent it having watched it not have a band indicator for more than 30 seconds.  Hopefully you got it and maybe it'll tell you something you didn't already know about the issue.

3) I don't yet have a B71 phone, though once a cheap Motorola one comes out, I plan to replace my T-Mobile 700 phone with one.  What kind of work-around are you considering?  Perhaps it could make sense to allow people to select which bands their phone actually supports or which bands they expect to see so you could work around obviously-incorrect data, even if it were in some kind of hidden menu.  (In my ideal world, there would be a hidden menu with lots of advanced functions to tweak the behavior of SCP in ways the average user may find unhelpful or confusing.  Like forcing all LTE to be identified as LTE 700 on my phone that is locked to 700.)

4) Is there any chance you could add a third mode to the neighbor cell notes beyond matching to PLMN and not matching to PLMN?  I'd like to merge my four SCP databases together (Verizon, AT&T, T-Mobile, and Sprint/USCC) but if I do that, then on my Sprint/USCC phones with the PLMN matching disabled, I'll get a bunch of neighbor cell notes that are clearly not correct since they'll match non-Sprint values.  And, of course, if I enable PLMN matching, then if I'm on Sprint, I'll miss Clear notes, etc.  It would be nice if there were a "Sprint" mode that matched only to PLMNs associated with Sprint when connected to a Sprint-associated network.  For example, if connected to PLMN 310120, have it match also PCI values for 311870 and other Clear PLMNs, but not to other PLMNs.  Or if connected to 311870, match to 310120, etc.

- Trip

[ingenium]

Yes, on a normal 8T8R site, 03/04/05 is the normal second carrier. On 4T4R tri-band antennas in Samsung markets, 09/0A/0B has been used fairly regularly. I do not know off the top of my head what is done with the 3rd carrier on the tri-band antenna sites (normal 8T8R being 06/07/08, which is already covered).

Third carrier on triband sites (that use triband endings) has second carrier GCI endings. So:

Non-triband:
2nd carrier - 03, 04, 05
3rd carrier - 06, 07, 08

Triband:
2nd carrier - 09, 0A, 0B
3rd carrier - 03, 04, 05

I define second carrier as the middle earfcn, and third carrier as the highest earfcn.

Sent from my Pixel 2 XL using Tapatalk

[mikejeep]

On 4/11/2018 at 9:14 PM, Trip said:

1) As indicated previously, in my area (DC) which I believe is a Samsung Clear market, 03/04/05 are third carrier, but SCP currently doesn't note it as anything other than standard B41.

2) Yesterday, I finally managed to send you diagnostics showing "LTE" without a band indicator while connected to a Clear site.  I sent it having watched it not have a band indicator for more than 30 seconds.  Hopefully you got it and maybe it'll tell you something you didn't already know about the issue.

03/04/05 are coded as third carrier, and your report showed sector 00 which should have showed you regular B41. You are likely getting victimized by the bug that has been driving me insane for a couple of years where things go south when the PLMN changes on the fly. Some devices suffer more than others. Essentially the band stops appearing until the PLMN changes again.. something is getting hung up momentarily on the OS side, and SCP chokes on it. It's related to the glitch that causes the PLMN and GCI to fall out of sync when switching cells, which also wreaks havoc and creates duplicate (bogus) database entries. It's maddening and I have tried all sorts of hacks i in the background to address it -- and I will continue to try, because it's a priority.

 

On 4/11/2018 at 9:14 PM, Trip said:

3) I don't yet have a B71 phone, though once a cheap Motorola one comes out, I plan to replace my T-Mobile 700 phone with one.  What kind of work-around are you considering?  Perhaps it could make sense to allow people to select which bands their phone actually supports or which bands they expect to see so you could work around obviously-incorrect data, even if it were in some kind of hidden menu.  (In my ideal world, there would be a hidden menu with lots of advanced functions to tweak the behavior of SCP in ways the average user may find unhelpful or confusing.  Like forcing all LTE to be identified as LTE 700 on my phone that is locked to 700.)

I'm hoping to keep it simple.. essentially by checking if it's a T-Mobile connection, and if so, checking if it's a band that T-Mobile actually uses. If not, assume it's a glitch and add 65536 to the EARFCN. I've spent the past few weeks trying to bring SignalCheck Lite up to speed and I have to finish that first, but it's not going well since it's been 3 years since that was updated and a lot has changed since then.

 

On 4/11/2018 at 9:14 PM, Trip said:

4) Is there any chance you could add a third mode to the neighbor cell notes beyond matching to PLMN and not matching to PLMN?  I'd like to merge my four SCP databases together (Verizon, AT&T, T-Mobile, and Sprint/USCC) but if I do that, then on my Sprint/USCC phones with the PLMN matching disabled, I'll get a bunch of neighbor cell notes that are clearly not correct since they'll match non-Sprint values.  And, of course, if I enable PLMN matching, then if I'm on Sprint, I'll miss Clear notes, etc.  It would be nice if there were a "Sprint" mode that matched only to PLMNs associated with Sprint when connected to a Sprint-associated network.  For example, if connected to PLMN 310120, have it match also PCI values for 311870 and other Clear PLMNs, but not to other PLMNs.  Or if connected to 311870, match to 310120, etc.

Hmm, that's an interesting thought. It would only make sense for Sprint, since most other major providers stick with one PLMN.. I'll look into seeing how feasible that is. What device do you see Clear neighbors when connected to Sprint, and vice versa? My Pixel doesn't do that.

-Mike

[Trip]

On 4/15/2018 at 8:31 PM, mikejeep said:

03/04/05 are coded as third carrier, and your report showed sector 00 which should have showed you regular B41. You are likely getting victimized by the bug that has been driving me insane for a couple of years where things go south when the PLMN changes on the fly. Some devices suffer more than others. Essentially the band stops appearing until the PLMN changes again.. something is getting hung up momentarily on the OS side, and SCP chokes on it. It's related to the glitch that causes the PLMN and GCI to fall out of sync when switching cells, which also wreaks havoc and creates duplicate (bogus) database entries. It's maddening and I have tried all sorts of hacks i in the background to address it -- and I will continue to try, because it's a priority.

 

I'm hoping to keep it simple.. essentially by checking if it's a T-Mobile connection, and if so, checking if it's a band that T-Mobile actually uses. If not, assume it's a glitch and add 65536 to the EARFCN. I've spent the past few weeks trying to bring SignalCheck Lite up to speed and I have to finish that first, but it's not going well since it's been 3 years since that was updated and a lot has changed since then.

 

Hmm, that's an interesting thought. It would only make sense for Sprint, since most other major providers stick with one PLMN.. I'll look into seeing how feasible that is. What device do you see Clear neighbors when connected to Sprint, and vice versa? My Pixel doesn't do that.

-Mike

Mike,

I didn't mean to conflate the first two items.  In the first, when I'm connected to Clear B41 third carrier on sectors 03/04/05, I'm not seeing the "3" next to it.  That's what I was trying to say.  I would love if this could be corrected, since it's hopefully a small change.  I'll try to send you a diagnostic showing it, but here's a picture from my log:  https://imgur.com/a/7zw72

I separately sent you the data on the missing band indicator because you said it was causing you a problem that you were having trouble tracking down, but I had been unable to send you diagnostic data on it previously.  I finally tried and successfully sent you the data without getting an error message.  I was hoping it helped you track down the issue, as it was sitting there for a very long time with no band indicator.  Same thing happened yesterday, actually, on a different site.  In that case, it lasted for a long time, measured in minutes.

Sounds like an interesting fix, though I suspect you'll have some Band 66/Band 2 overlap.  For example, if T-Mobile held AWS-1 A-block, that would be EARFCN 66586, and 66586-65536=1050, a valid Band 2 EARFCN.  Not sure what you would do about that.

Yes, I agree it would have to be Sprint-only.  I think most of the phones I've had since moving to Sprint would show Clear and Sprint B41 in neighbor cells at the same time.  (LG G6, LG G5, Samsung S5.)  They do/did not, however, show PCI values for US Cellular or other networks.  I'll confirm and let you know for certain.

- Trip

[Trip]

1 hour ago, Trip said:

Yes, I agree it would have to be Sprint-only.  I think most of the phones I've had since moving to Sprint would show Clear and Sprint B41 in neighbor cells at the same time.  (LG G6, LG G5, Samsung S5.)  They do/did not, however, show PCI values for US Cellular or other networks.  I'll confirm and let you know for certain.

Mike,

Confirmed via picture.  https://imgur.com/a/ZWewf  I'm connected to a Clear site with two 8T8R sites appearing in the neighbor cells.  Couldn't get diagnostics before it dropped back to B26 on the DAS though.  I'm in too much pain to walk back to the window and try again for diagnostics.

- Trip

[bigsnake49]

On 4/11/2018 at 8:22 PM, mikejeep said:

I have received a lot of messages over the past 3 weeks regarding improper LTE band identification for T-Mobile users. There is an Android bug that causes EARFCNs assigned to newer LTE bands (>46) to be reported incorrectly, such as Band 66 and Band 71. SignalCheck is showing Band 2/3/5/6/7/etc. because that is what the OS is reporting.

For more details, please see this bug report; I encourage everyone to 'star' it and/or adding your own report to try getting it more attention: https://issuetracker.google.com/issues/37136986

I am trying to implement a workaround that will at least help T-Mobile users until Android fixes the issue.

-Mike

 

Yeah mine is reporting B3 for T-mobile. I am definitely in the US and it should be reporting B2.

Edited May 5, 2018 by bigsnake49

[greenbastard]

6 minutes ago, bigsnake49 said:

Yeah mine is reporting B3 for T-mobile. I am definitely in the US and it should be reporting B2.

You're probably getting Band 66 instead of Band 2 but see Band 3 instead.

[bigsnake49]

4 minutes ago, greenbastard said:

You're probably getting Band 66 instead of Band 2 but see Band 3 instead.

Very possibly! My phone supports Band 66.

[Jesito473]

My phone when it attaches to B66 here in Cincinnati, it shows up as B3 (1800mhz) as well. The only app that gives an accurate reading is Cellmapper at the moment.

Very possibly! My phone supports Band 66.

Sent from my SM-G965U using Tapatalk

[jefbal99]

Hey Mike, will this new tightening on network monitoring impact SCP?

https://www.androidpolice.com/2018/05/07/apps-android-p-will-no-longer-able-monitor-network-activity/

Apps on Android P will no longer be able to monitor your network activity

[Trip]

I won't speak for Mike, but it sounds like they're specifically talking about network traffic, as in what websites you're connecting to, rather than which cell tower you're connected to.  That information comes from an actual Android API which requests permissions, and what they're talking about here is direct access to /proc/net in the file system, which apparently does not.

That said, it's vague enough that I, too, would like to hear more information on it. 

- Trip

[mikejeep]

9 hours ago, jefbal99 said:

Hey Mike, will this new tightening on network monitoring impact SCP?

https://www.androidpolice.com/2018/05/07/apps-android-p-will-no-longer-able-monitor-network-activity/

Apps on Android P will no longer be able to monitor your network activity

I saw that article as well and I probably turned white as a ghost for a few moments of panic.. but after reading it, and poking around a bit more, I do not think this will impact SCP at all. As Trip stated it looks like it's closing off direct access to /proc/net, which isn't anything I use. But, we won't be certain until P gets closer to release.

-Mike

[bakedc4]

On my Essential running Android P. Seems to have some extra info.

 

 

Sent from my PH-1 using Tapatalk

 

 

 

 

[Dkoellerwx]

9 minutes ago, bakedc4 said:

On my Essential running Android P. Seems to have some extra info.

 

Sent from my PH-1 using Tapatalk

That looks like the normal amount of info?

[bakedc4]

That looks like the normal amount of info?

Unless I'm having serious memory issues (entirely possible, way too much work and not enough sleep) the neighbor cells only showed up if they were in the same band and all others showed as unknown on 8.1

Sent from my PH-1 using Tapatalk

[Dkoellerwx]

1 minute ago, bakedc4 said:

Unless I'm having serious memory issues (entirely possible, way too much work and not enough sleep) the neighbor cells only showed up if they were in the same band and all others showed as unknown on 8.1

Sent from my PH-1 using Tapatalk
 

Ah, depends on device. Some changes have been made that now shows other bands in the neighboring cells, even for devices on Oreo. 

[BlueAngel]

2 hours ago, Dkoellerwx said:

Ah, depends on device. Some changes have been made that now shows other bands in the neighboring cells, even for devices on Oreo. 

Yeah I am seeing two B41 and one B25 on my neighboring cells S9+ here.