South of Cambridge used to get 4G now 2G for days

[johnnybaloney]

Hello,

I used to consistently and reliably get 4G in the south of Cambridge but recently (for at least a week) only get 2G. The satnav struggles, opening websites is slow etc. and I see the dreaded 'E' by the signal strength indicator. Any idea what's happening?

[johnnybaloney]

The log above is from the phone with the EE sim, I had both phones on me. I thought it didn't matter which side initiated the connection and if I had 4G reception I'd get the call on 4G. I was planning to drive into my area to test the handover but the call arrived on 2G. I can try callng from the EE phone instead if that would make it use VoLTE.

[bristolian]

It's irrelevant whether the call is inbound or outbound, what matters is that a call is started on VoLTE and you travel into the footprint of the affected site.

Cellmapper will help you verify that you are attached to the offending site (make a note of the eNB-ID, PCI & EARFCN to pass to CS),  The purpose of the exercise is to identify whether the site takes incoming VoLTE handovers.

Your investigations so far seem to suggest that phones in idle on this site - potentially only on certain carriers - are being pushed down to 2G. You will ultimately need to pursue this with CS (good luck!!) but these checks may assist them.

[johnnybaloney]

Unfortunately I'm unable to perform this test. As soon as I start a call or receive a call while in 4G reception area I immediately get switched to 2G for the duration of the call.

I received a call from the tech level 2 team today but they couldn't immediately identify any fault anywhere. Will call again tomorrow.

[bristolian]

The response from "Level 2 tech" doesn't surprise me.

If call attempts from a 4G-idle phone are dropping to 2G, this indicates CSFB rather than VoLTE, is in use. If your phone is doing this in multiple locations (or on multiple sites, to be exact) this indicates a device/SIM issue.

If your phone is only doing this in particular locations, it suggests a config issue on that site/sector. I'd be minded to ask the "Level 2" team to check the VoLTE success/failure metrics. Your end result should be a fault ticket from CS (I include "Level 2" in the CS bracket) into the networks area, for further checks.

[johnnybaloney]

An interesting development. After receiving an email update with the following:

---

---

---

Network update: We could not find a problem in CB22 3BY. This may have been a one off and we therefore recommend you restart your device

---

(not my postcode, not sure where it came from, but close enough) I decided to give 150 a call again and request a replacement sim card. Given that everything else was reportedly working fine it remained one of few options left for me to try.

I described the problem again, mentioned that I tried the sim with yet another phone with the same results, also mentioned about the EE status checker that has never worked for me (still getting "Sorry, there's a problem with the service. We're working to fix the issue.").

Then the technician made a discovery that there was some work ongoing on a mast near M11 that's been causing trouble to 18,000 households for the last two weeks! The boundary of the impacted area matched exactly with what I experienced while on my test route described in the earlier post where I could get to a place with 4G signal. My postcode was not marked as affected in their system as I'm on the fringe of the affected area.

So we will see. The work on the mast is scheduled to finish end of day tomorrow.

[bristolian]

This is where Cellmapper comes into its own. You can use its (usually pretty accurate) location data to prove your serving site at any given time.

You can then prove or disprove the "technician" theory at a stroke. Given that you can explain clearly what the symptoms are (lack of VoLTE, CSFB to 2G on call attempts), were they able to be specific on what the "trouble" was?

It's most unusual for a network to actively encourage call attempts down to 2G, all performance criteria & customer-experience metrics would be for the exact opposite.

[johnnybaloney]

I'm not sure they are too keen on engaging in discussion at this sort of level. It's probably not too controversial for them to assume that the caller probably has little clue about the technical matters. The previous technician did say that he was using cellmapper outside hours but they had their own systems and data so essentially wouldn't comment on what I could see in cellmapper logs.

Today the technician wasn't too specific on the nature of the problem with the mast beyond mentioning 2G, 4G, affected spots and number of households.

[bristolian]

---

@johnnybaloney wrote:

I'm not sure they are too keen on engaging in discussion at this sort of level. It's probably not too controversial for them to assume that the caller probably has little clue about the technical matters.

---

CS are indeed geared up for the majority of customers whom don't have any interest in how their phones work, they just want it to work - their script-based processes indeed follow this logic.

[johnnybaloney]

An update on the situation. It appears that the problem has been fixed. For a few days I've been getting 4G without issue, no dropping to 2G any more.

It seems that the mast that was borked and the technician referred to was the one I mentioned earlier in the thread with eNB ID 21548. I'm guessing the repair work on it was completed and it's now back to full operation. If that's the one that is because on Twitter I was told at the time:

---

We don't have any logs of faults on it right now but I can see a change in its performance.

---

I wish those affected by the mast issue were notified about the work on it. How hard is it to send a letter or even a text to those living within the affected area. Trying to get to the bottom of the problem was a real PITA.

I would like to thank @bristolian for helping me here. I was seriously considering ditching EE if not for their assistance leading to finding the root cause of the problem.

[johnnybaloney]

P.S. The service status checker still doesn't work for me...

This is what arrives from the backend:

---

{"code":null,"message":"Bad Request"}

---

They still "are working to fix the issue" 🙄