Likely not a TTN/LoRaWAN use case, unless you can do some processing at the nodes to consolidate and extract “information” vs “data” and send a processed result less frequently - say ever 20 mins with similar payload size.
What data are you measuring that changes every 500ms that you need to capture?!!!
If it helps your thinking see:
https://avbentem.github.io/airtime-calculator/ttn/eu868/20
So is that 32 nodes for each of the 100 sections ?
And is it required that all sent data is acknowledged, so that the node is sure its been received ?
That is not a LoRaWAN use case in my opinion. Too much traffic and certainly resulting in significant data loss. May-be you can use other Lora based technology however that I know nothing about.
With regards to other options, I don’t have a fitting answer at hand. You could ask Mouser product specialists for input. Or another company, there are plenty specializing in RF connectivity. Mouser and companies like it have no vested interest in a specific technology and should be able to provide objective advise.
This rate is for peak times. Maybe four or five times a day, but all transceivers must work together or with a very tiny latency.
At peak times LoRaWAN can’t do 500ms, but I would like to know what parameter in Agri changes that fast?
What does ‘work together’ mean ?
What exactly is a ‘very tiny latency’ ?
You might know the answers to those statements, but the forum cannot even guess.
Sounds like the application is not possible within TTN and may even exceed legal duty cycle limits.
This! Apart from the whole setup & tx & rx1 & rx2 timings, there’s the inexplicable change that requires that such short timings and in parallel, something almost all radio systems would struggle with as they compete for the gateway’s attention.
@salimi, scroll up to the top of this page, the Learn link will explain many of the challenges you face applying LoRaWAN to this scenario & may explain enough basic theory for the other radio systems that would struggle.
If we had more info on the what & why it may be possible to suggest a solution - like recording the data but having the uplinks staggered.