On some LoRaWAN Gateways (e.g. Dragino LPS8N) the admin setup offers separate connectivity options for Basic Station and for Semtech UDP. I’ve successfully just set up a gateway using the Semtec UDP setup and it was no big deal, but I’m interested in the advice of the community regarding this dichotomy of options.
Are both of these setup options supported on TTNv3?
What are the advantages / disadvantages of each?
Which one should you use given the choice?
I’m not sure if the instructions online are just dated, but the TTI instructions for the LPS8N seem to only cover setup using the Semtech UDP mechanism. Is that intentional? Here’s a look at the Basic Station setup page for the LPS8N, where would you get the info needed to fill in these fields?
My guess is on the TTN console page for the registered Gateway. Can you change this on the TTN console after the fact, or do you need delete and re-register the Gateway to do this? Here’s what I see on the Gateway settings page on TTN Console:
UDP been around since LoRaWAN start, uses UDP protocol (Dho!) unroutable and essentially fire and forget - basically just works, or may have problems if backhaul sevice provider screws with UDP messages. Its Simple and doesnt require security keys etc. Supported by many LNS Services and by many GW manufacturers, generally well understood. However, considered not as secure as…
Basic Station uses secured IP connection, uses routable traffic, allows retries (but may be of limited value in high latency environments as retry may fall outside acceptance window and fail anyhow). Needs a secure key hence the Authentication Key option above. As such harder to spoof implementation.
In a fully secured production environment you might therefore steer to BS, however, in reality in TTN context as a experimental sandbox and community environment its probably a ‘dont care’ and for many users who are not into the details of secure networking infrastructure its an extra complication to worry about. From your nodes broadcast traffic perspective again it’s a ‘dont care’ as the broadcast will be heard by all GW’s in range irrespective of PF/traffic type and will be sent on to the LNS for de-duplication so even if you have BS GW its possible probably many others will be sending over UDP so . Older GW’s tend to be UDP hence most TTN GW’s (and many on peered networks that also feed into TTN) will use that. From approx 50 GW’s in ‘personal’ fleet I would say 1/3+ (maybe ~50%) on UDP, 1/4+ on BS, and rest ‘other’ if that is any help!
Are you asking if Semtech UDP and Basic Station are supported on TTS? If so, this is clearly shown in Concepts | The Things Stack for LoRaWAN. If you mean the LPS8N, if it supports the two standards, then, well, then by being standards that are supported by TTS, the answer would be yes.
I’d use an SDR + Avian over IP … … there is nothing useful in asking people what they’d use, it’s down to your own use case and/or the hardware - getting a TTIG to do Semtech UDP would be silly and if I wanted to secure my already encrypted payloads, I’d use BS. I think the primary benefit overall is that BS isn’t UDP so less likely to get lost in transit, for the very minute values of lost packets.
The intent is rather hard for us to determine, as that would be in the heads of the TTI docs team. But if you do a forum search, you’ll find a number of people using BS with the LPS8N. Most likely its just a case of firmware moving on but not being drawn to the attention of those creating the docs, I don’t think there is anything sinister going on here.
Why can’t you try this yourself? Even by setting up a fake gateway on UDP and then trying to change it?
If you do want to change, remember the golden rule of searching the forum for the rather overused but under searched phrase “EUI is not ID”
. Older GW’s tend to be UDP hence most TTN GW’s (and many on peered networks that also feed into TTN) will use that. From approx 50 GW’s in ‘personal’ fleet I would say 1/3+ (maybe ~50%) on UDP, 1/4+ on BS, and rest ‘other’ if that is any help!