Rogue nodes flooding a gateway

Hi All,

I have a gateway on a farm that is getting hammered by some unknown nodes.

The cnt’s are all at 0 and they all have this Net ID

image

Does anybody know what might be happening here?

Getting one through a second across about 10 rogue nodes

Worried about air time being swamped - data usage of the gateway hasn’t gone up any significant amount, is that because non-TTN packets are discarded?

cheers
Paul

How far away are you from other gateways?

15-20km to the nearest TTN gateway.

I am thinking a farm might have been setup recently with some automation hardware that uses Lora

yea maybe… I wouldn’t be too worried. Someone maybe just getting the feel for it and are testing. If it goes on and on then I would be concerned. I hardly see any traffic through my gateway and am looking forward to farmers around me picking up the tech and using it.

Those probably aren’t TTN; they may not even be LoRaWAN

Could you post some example raw packets with SF, frequency, and radio payload?

1 Like

This is Australia

Here is the payload
402D4B480180000001AA050C42476898

Dev Address
01484B2D

  {
      "gw_id": "ttn_stgeorge",
      "payload": "QC1LSAGAAAABqgUMQkdomA==",
      "lora": {
        "spreading_factor": 10,
        "bandwidth": 125,
        "air_time": 329728000
      },
      "coding_rate": "4/5",
      "timestamp": "2020-12-22T09:11:41.467Z",
      "rssi": -121,
      "snr": -8.75,
      "dev_addr": "01484B2D",
      "frequency": 917800000
    }

If you want to see anything else, let me know

Locals have been doing some enquiries - seems like a company has been setting up some irrigation automation, using their own gateways.

30 nodes sending at a minimum each 2 minutes. Still confirming if the dev address is their nodes or not.

If so, interesting decision for a commercial company to use the same sub band as the largest public network in Australia (sub band 2) - i thought it would best for all to use a different sub band

Seems a very high update rate under any circumstances but particularly for shared frequencies. It’s not like the soil is going to change dramatically in 2 minutes. It wouldn’t be so hard for them to only send on change of values rather than a continuous stream.

It may be worth talking to the company as it may be that most of the staff aren’t aware of the impact of a decision that someone technical took based on “because we can” rather on need. Or marketing like the idea of 2 minute updates.

It is for an irrigation network, so I think the 2 minute updates are more for receiving downlinks rather than uplinks, to be responsive as required.

Class C devices would make better sense

Absolutely - as would some on-device intelligence so it didn’t need the 2 minute cycle even at Class C.

You have to wonder what would happen if something came along that flooded the airwaves and their devices could no longer be commanded in a timely fashion.

Could go up to the farm, make contact and see if they are implementing something. As techies go, we always like to show off our stuff when people pay an interest. Just notice you’re in OZ. So visiting the farm maybe be a bit of a day trip.

Yes 14 hr round trip :grin: