I work as a telecom engineer troubleshooting/programming the Ciena 5164 optical router used for Verizon and At&t. These companies and many of my colleagues know exactly what it was but the more interesting part that raised some eyebrows was the fact that cc payment processing and emergency services/SOS/911 calls WERE possible for most of the affected customers. This is key. (This all runs on the same fiber optic network and should have been ‘down’ also.)
I cannot overstate how anomalous the above fact pattern is because when our networks go down, they don’t go down in this piecemeal way.
Here’s the takeaway: The goal of a ‘cyber weapon’ is maximum disruption. Whoever deployed (tested) this ‘cyber tool’ went to great lengths to pause communication ONLY. This surgeon scalpel approach we saw was a MUCH more difficult task than to just ‘nuke’ the whole system so to speak. A tool like this would need to be tested a few times to ensure a reliable partitioning of the ‘Target’ [calls/data] and the ‘Not Target’ [911/cc processing].
I'm a patriot who loves their country so I will not say what 'it' was/is, but if you reread the above a few times, and ask yourself the right questions, logical deduction should provide the answer.
Are most businesses really using cellular networks for their CC processing? Not too long ago I know that satellite links were popular for that application. And there are MVNOs that can operate on multiple carriers. That sounds like a good idea for a high reliability service like CC processing. Likewise emergency calls can go through any available network. I haven't heard anything suggesting tmobile was also affected.
I know this is not what you are implying and it's late so chances are no one will read this, but here goes…. Modern cellular networks require precise time synchronization, which is typically provided by a GPS disciplined oscillator. These oscillators have a holdover specification, a time period for which it can maintain accurate enough time for the system to fully function without a GPS signal. Now, potentially the networks should be designed to operate without such precise synchronization. But in this case the network capacity would be greatly reduced. If that happened, it would make sense to only serve the most important applications.
Calls, credit card processing, 911/SOS, streaming/Internet… basically any type of civilian or commercial communication is essentially DATA and travels along our fiber optic network in the form of light pulses. We called this traffic.
Our fiber optic network can be compared to our highway system and the data that traverses is it, traffic on that highway system.
There is no way for this fiber optic highway to "go down" the way it did. This is because it didn't "go down"the way they would normally go down for a cyber attack or software glitch.
It just got really picky for some non-accidental reason. Impossible that it wasn't deliberate based on the fundamentals of how the network moves data.
Here's the bottom line: Someone went out of their way to engineer the capability to pause communication without disrupting society. Now they are working the bugs out. When upgrading existing networks, half my time is spent troubleshooting until the system works properly. Nothing in this industry works correctly the moment you hit the return key. Nothing. There is always bugs to work out.
I wouldn't be surprised if we see several more tests over the near future before the main event(s). I suspect the outages will be described with more and more uniform/consistent fact, patterns as the bugs are worked out of the tool.
While you're thinking on the above, think about this…
"Internet will be paused." "Blackout necessary." (comms) "It must be hard to communicate." ~Q