Jump to content

SignalCheck Beta Crew Forum


mikejeep

Recommended Posts

On 12/30/2021 at 5:05 PM, PedroDaGr8 said:

Hi Mike, a question about gNBs to which you don't connect? I noticed if I band lock my phone to n5 or n66, I can see AT&T n5-SA and n66-SA bands. I can't connect to them but I can see them in CellMapper and in SignalCheck. That being said the IDs are different, the CellMapper gNB appears to match the eNBs in the area a bit closer than SignalCheck does (the first digit is the same).

Additionally, the Signal check results don't upload while the CellMapper results do.

 

Ignore what I said about the gNB, yours seems to be more consistent than CellMapper when it comes to the Sector IDs. My guess is they don't have the right number of bits set for the gNB. That being said, the sites still don't upload.

  • Like 1
Link to comment
Share on other sites

On 12/31/2021 at 7:26 PM, PedroDaGr8 said:

Ignore what I said about the gNB, yours seems to be more consistent than CellMapper when it comes to the Sector IDs. My guess is they don't have the right number of bits set for the gNB. That being said, the sites still don't upload.

When you say don’t upload, what do you mean? Can you send me the file the app spits out on failure?

Link to comment
Share on other sites

19 hours ago, RAvirani said:

When you say don’t upload, what do you mean? Can you send me the file the app spits out on failure?

There's no failure or error. Based on my testing today it appears that it never records the cells. I tested it today while driving around and despite seeing a dozen cells, nothing was recorded.

Link to comment
Share on other sites

9 minutes ago, PedroDaGr8 said:

There's no failure or error. Based on my testing today it appears that it never records the cells. I tested it today while driving around and despite seeing a dozen cells, nothing was recorded.

Oh gotcha - that one is a Mike question.

Link to comment
Share on other sites

On 12/30/2021 at 8:05 PM, PedroDaGr8 said:

Hi Mike, a question about gNBs to which you don't connect? I noticed if I band lock my phone to n5 or n66, I can see AT&T n5-SA and n66-SA bands. I can't connect to them but I can see them in CellMapper and in SignalCheck.

Additionally, the Signal check results don't upload while the CellMapper results do.

On 1/2/2022 at 8:45 PM, PedroDaGr8 said:

There's no failure or error. Based on my testing today it appears that it never records the cells. I tested it today while driving around and despite seeing a dozen cells, nothing was recorded.

When you say you "can see" those bands but "can't connect" do you mean you see them as neighbor cells? I'm not completely sure I understand what you're interpreting as seeing vs. connecting..

Link to comment
Share on other sites

Another new SCP beta release is rolling out now! A full suite of 5G-NR icons with bands/frequencies are now available. A new status bar icon style has also been added. If you select "Band" the icon will simply show the connected band with  either an "n" prefix (5G-NR) or a "B" prefix (LTE). This provides a more visible option for users to see what band they are currently connected to, with less extra text causing smaller fonts to be needed.

Thanks!

Link to comment
Share on other sites

14 hours ago, mikejeep said:

When you say you "can see" those bands but "can't connect" do you mean you see them as neighbor cells? I'm not completely sure I understand what you're interpreting as seeing vs. connecting..

Not as a neighboring cell, the information fully populates in SignalCheck (see image below), I just don't have a data connection. 

9tTuYj3.jpg

Furthermore, despite driving around for a while and seeing multiple n5-SA cells, nothing is available for upload 

Also, I'm still investigating the gNB. I thought all the Cell IDs for SignalCheck we're consistent: 1049/1073/1097 for n5-SA, 1050/1074/1098 for n2-SA, and 1051/1075/1099 for n66-SA. Said above image also shows a two digit Cell ID. It appears this is also a set 25/49/73 which are n5-SA.

If it helps, device is a Samsung Galaxy Note 20 Ultra running Android 11 (Nov. Security). I sent you a diagnostic report a week or so ago while seeing an AT&T cell. 

Edit:

So I ran some calculations and it appears that CellMapper is using 24-bit gNB for AT&T, while SignalCheck is using a 25-bit gNB. Not sure which is correct yet since I can't find a screen to show me said info (my radio data doesn't populate since switching to a TMo sim).

Edit2: Actually, using more NCI data from CellMapper, I think both SignalCheck and CellMapper are calculating gNB wrong. Based on the NCIs I have access to, if you use 26-29bits to calculate the gNB then you get a very consistent pattern for the Cell IDs:

n5: 25/49/73

n2: 26/50/74

n66: 27/51/75

n78: 28/52/76

Link to comment
Share on other sites

8 minutes ago, RAvirani said:

T-Mobile is doing 2048 cells per gNB and AT&T is doing 1024 cells per gNB, I believe. 

I might need to double check that though. 

That 1024 number would be in line with 26-bits for the gNB. 10-bits = 1023 which including 0 equals 1024.

Link to comment
Share on other sites

17 hours ago, PedroDaGr8 said:

Not as a neighboring cell, the information fully populates in SignalCheck (see image below), I just don't have a data connection.

Furthermore, despite driving around for a while and seeing multiple n5-SA cells, nothing is available for upload 

That behavior is by design -- but I'm flexible. I'll get back to you on that. Sent you a PM about the gNB questions.

Link to comment
Share on other sites

12 hours ago, RAvirani said:

T-Mobile is doing 2048 cells per gNB and AT&T is doing 1024 cells per gNB, I believe. 

I might need to double check that though. 

Apologies - T-Mobile is doing 4096 cells per gNB (12 bits per local ID and 24 bits per gNB). 

Still working on AT&T…

  • Like 2
Link to comment
Share on other sites

Many thanks for the new status line icons.

 For comments above, when I take the global cell Id on the service mode screen *#0011# of this Samsung Galaxy A32 5g T-Mobile and convert it to hex I get the same as the NCI on SCP.

  • Like 1
Link to comment
Share on other sites

6 hours ago, RAvirani said:

Apologies - T-Mobile is doing 4096 cells per gNB (12 bits per local ID and 24 bits per gNB). 

 

Ouch, if true, that means that SCP isn't calculating the gNB correctly for T-Mobile either. So all of the gNB online are incorrect. For example the Woodinvile Warehouse Site would have a gNB of 1864948 rather than 3729896. 

Link to comment
Share on other sites

22 minutes ago, PedroDaGr8 said:

Ouch, if true, that means that SCP isn't calculating the gNB correctly for T-Mobile either. So all of the gNB online are incorrect. For example the Woodinvile Warehouse Site would have a gNB of 1864948 rather than 3729896. 

Not too severe of a problem, the actual raw numerical NCI value is what is recorded by SCP, what you see in the app  (and on the map) is that value with various math/formatting applied. Looks like it's already been adjusted on the map, at least for T-Mobile in my area. I'll be tweaking SCP based on what the consensus is for appropriate values for the major providers.

  • Like 2
Link to comment
Share on other sites

3 hours ago, PedroDaGr8 said:

Ouch, if true, that means that SCP isn't calculating the gNB correctly for T-Mobile either. So all of the gNB online are incorrect. For example the Woodinvile Warehouse Site would have a gNB of 1864948 rather than 3729896. 

Yup - both us and Cellmapper had it wrong.

As Mike said, we store the full NCI, not the gNB or sector, so recalculating these values is pretty easy for both of us. I’ve already done so on the map.

One of the red flags I’m surprised Cellmapper didn’t pick up on is that there are no odd gNBs when using a 25 bit gNB. 

  • Like 1
Link to comment
Share on other sites

Noticed when NSA is activated, the status bar displays the nXX as planned for T-Mobile using TNX Sim and AT&T using Boost Mobile black 5g sim with latest SCP, contrary to earlier (now corrected) report.

 

Edited by dkyeager
Link to comment
Share on other sites

On 1/10/2022 at 8:08 AM, mikejeep said:

Another SCP beta rolled out yesterday afternoon.. more bugfixes and improvements. NR NCI formats have been adjusted as discussed above.

FYI, when I exported site logs it still has the old NCI format.

  • Thanks 1
Link to comment
Share on other sites

A new SCP beta is rolling out now and should be available shortly! This update should fix the GCI and NCI issues on exported site logs mentioned above, as well as some cleanup of unused log columns and tables. I also changed the method the app uses to identify and report connections; when the device is reporting it is out of service but still seeing cell info, it should now still record that cell info in the log. This will need some testing; issues may arise while roaming or in fringe signal areas, but we will see.

If this proves to be stable over the next several days, it will become a public release next week. Fingers crossed!

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

18 hours ago, mikejeep said:

I also changed the method the app uses to identify and report connections; when the device is reporting it is out of service but still seeing cell info, it should now still record that cell info in the log. This will need some testing; issues may arise while roaming or in fringe signal areas, but we will see.

On 12/30/2021 at 5:05 PM, PedroDaGr8 said:

Hi Mike, a question about gNBs to which you don't connect? I noticed if I band lock my phone to n5 or n66, I can see AT&T n5-SA and n66-SA bands. I can't connect to them but I can see them in CellMapper and in SignalCheck. That being said the IDs are different, the CellMapper gNB appears to match the eNBs in the area a bit closer than SignalCheck does (the first digit is the same).

Additionally, the Signal check results don't upload while the CellMapper results do.

Excited to see if AT&T SA NR data will upload now!

  • Like 2
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • large.unreadcontent.png.6ef00db54e758d06

  • gallery_1_23_9202.png

  • Similar Content

  • Posts

    • A heavy n41 overlay as an acquisition condition would be a win for customers, and eventually a win for T-Mobile as that might be enough to preclude VZW/AT&T adding C-Band for FWA due to spreading the market too thinly (which means T-Mobile would just have local WISPs/wireline ISPs as competition). USCC spacing (which is likely for contiguous 700 MHz LTE coverage in rural areas) isn't going to be enough for contiguous n41 anyway, and I doubt they'll densify enough to get there.
    • Boost Infinite with a rainbow SIM (you can get it SIM-only) is the cheapest way, at $25/mo, to my knowledge; the cheaper Boost Mobile plans don't run on Dish native. Check Phonescoop for n70 support on a given phone; the Moto G 5G from last year may be the cheapest unlocked phone with n70 though data speeds aren't as good as something with an X70 or better modem.
    • Continuing the USCC discussion, if T-Mobile does a full equipment swap at all of USCC's sites, which they probably will for vendor consistency, and if they include 2.5 on all of those sites, which they probably will as they definitely have economies of scale on the base stations, that'll represent a massive capacity increase in those areas over what USCC had, and maybe a coverage increase since n71 will get deployed everywhere and B71 will get deployed any time T-Mobile has at least 25x25, and maybe where they have 20x20. Assuming this deal goes through (I'm betting it does), I figure I'll see contiguous coverage in the area of southern IL where I was attempting to roam on USCC the last time I was there, though it might be late next year before that switchover happens.
    • Forgot to post this, but a few weeks ago I got to visit these small cells myself! They're spread around Grant park and the surrounding areas, but unfortunately none of the mmwave cells made it outside of the parks along the lake into the rest of downtown. I did spot some n41 small cells around downtown, but they seemed to be older deployments limited to 100mhz and performed poorly.    
    • What is the cheapest way to try Dish's wireless network?  Over the past year I've seen them add their equipment to just about every cell site here, I'm assuming just go through Boost's website?  What phones are Dish native?  
  • Recently Browsing

    • No registered users viewing this page.
×
×
  • Create New...