I saw that these are shipping and have almost convinced myself to buy one. I think they're pretty neat and now a wideband phased-array direction finder is a project that a smart high-schooler can do, at a price point that many people can reach. Maybe the next time this type of interference occurs the headline will read different: "Local kid solves problem with science fair project, stumped adults for weeks."

With that said, these KerberosSDRs (or any multi-channel coherent SDR solution) are not widespread, are considerably more expensive (6-7x more than a single RTL-SDR), and are only recently available. So my previous point still stands about how armchair hackers can talk about how easy this problem is to solve in theory but they lack the actual equipment to do it.

Oh absolutely. A simple RTL-SDR is widespread, but anything coherent is absolutely not. I mean, the Ettus 4x4 pcie card is coherent, but it's also starting at $5000 for the card. And since it's a pcie, it goes in a very non-portable computer.

The KerberosSDR has one very significant disadvantage: the databus is usb2. The theoretical max is 30M samples at 8bit I and Q... And frankly we need much more bandwidth. But for a beginner foray into coherent radio applications, a KerberosSDR for $100~ is crazy worth it.

Now, the older way of doing RDF (I'm a Ham, btw), was to get a handheld receiver capable of listening to the freq in question, highly directional antenna, take measurements and move. Effectively, the human with a receiver is the manual triangulator which can be trained with experience. That was because gear that did triangulation was $$$$$$.