The most expensive skill our industry quietly stopped teaching is reading the air interface.
Not RF theory. Not dashboards. The actual messages.
Here is how it plays out, and you have all seen it.
A customer reports a cut. Someone opens the OSS. One radio link failure that hour, on that cell. The ticket goes to the vendor. The vendor asks for evidence. Two weeks later everyone agrees the cell “looks fine” and the ticket is closed as not reproducible.
Now the same event, from the device.
37 291 events in the capture. 207 flagged critical, every one of them a failure or a drop.
And the shape:
drop at 01:53, 01:56, 01:59. A pause. 02:42, 02:57. The radio link failure lands at 03:02. The drops keep going after it.
Seventy seconds of degradation with a signature, before the thing your counter recorded. The counter never had that. It could not have.
One more, because this one bothers me more than it should.
Open a measurement report and you find rsrpResult : 28. Twenty eight. That is the value the UE actually sent. Minus 113 dBm is the conversion, per TS 36.133. Most engineers I meet have never seen the left side of that arrow, only the right side, printed by a tool that decided for them.
So here is my position, and it is not a comfortable one.
If your team cannot open an RRC message, you do not have a troubleshooting capability. You have a reporting capability. Those are different budgets, different people, and very different outcomes when something breaks.
This is not a talent problem. Tools abstracted the air interface to sell simplicity, training budgets followed, and within one generation the layer became optional, then invisible, then unreachable.
I am not above this. I re-run my own fundamentals every six months, not to prove anything, to catch my own drift.
Question, and please be honest.
In your team today, how many people could open a capture and tell you what happened, without a dashboard?
LinkedIn: 