Would it be possible to cache known devices at the gateway bridge?

I am thinking of denial of service type attacks. If someone were to bring a bunch of devices to a location with a gateway and send packets continuously, could they bring down chirpstack components by generating traffic / errors?

If that’s a possible issue, is there a way to cache known devices at the gateway bridge?

As a start, the Gateway Bridge can currently filter by NetID and JoinEUI. That takes place after a packet has been received by the radio, of course, so it won’t help with saturated airwaves.

2 Likes

Yes and no. There are some papers handling collisions when too many nodes are in the same area. I dont know if the amount of messages from one gateway is enough to bring down a load-balanced “Chirpstack” (we are testing that on the second half of the year - also with jamming) - I think the gateway itself is going to collapse first.
A wrongly configured network, could lead you to a point where (at least in Europe) you run out of “Air Time” for the gateway (they also have to follow and respect the Duty Cycle) - but that is not a “classical DoS”
Is not as easy as dropping packets, it’s easier with cables - lots of points where you can choke the unwanted traffic. When dealing with radio waves, well…there is not much you can do besides going out with a zapper and “de-activating” the “devices” (jammers)