Off the the top of my head, I'd say it's doable on nRF52. Other questions would that would guide one in selecting hardware:
Do you want to forward packets from LoRA, BLE, or both?
Are you connecting the gateway to the internet via Wi-Fi or Ethernet?
Do you want single-channel, or multi-channel forwarding?
Where is the gateway to be installed?
The current status of TTN kickstart GW is no known delivery date which means I have to go to plan B: the "standard" Pi / ic880 : ie multi-channel.
However, the Pi can be termed overkill (opinionated).
So step 1 is get backing for the gw's: 4 to cover https://www.thethingsnetwork.org/community/silkeborg/ around the hills of the valley that is Silkeborg (history started with water and the production of paper).
step 2 is make one GW with Pi for presence
step 3 is a bit of play and learn (nRF52 (puck.js) instead of Pi) with some tests: limited time as a constraint so that may not happen until more time and extra ic880
step 4 with regard to time constraint, use what I have gotten so far for the other 3 locations (my house is location 1 as it is has height and internet with clear sight over Silkeborg to the other locations)
The Pi has everything, including the kitchen sink and an SD card one has to be careful with in connection with corruption with power loss (solved by R/O SD access). It has IP connectivity (Ethernet and wifi).
The nRF52 /puck.js does not have IP connectivity "out of the box" so that is a factor that means that there would be a need for an "extra" contra ESP8266 (Gordon has Espruino WiFi: I just have not has an excuse/wish to use it yet).
The answer is, "it depends" I know: is it "play and learn", probably.
(off to a meeting today to see if I can get backing)
Looking at the LoRa stuff, it sure seems like you could run a packet forwarder on Espruino.
nRF52 could easily forward BLE data over an RN2483 to a LoRa gateway, and could do so while using relatively little power.
Trying to make a gateway itself could be possible too, maybe - but I see that being a lot more work, and personally I think it'd be worth leveraging the existing work done on the Raspberry Pi version.
"Stuff" is on it's way: I tend to buy from people and not companies.
Pi will be production: nRF52 is going to be interesting.
Azzy'S Electronics Questions or problems?
The seller, DrAzzy, will ship your order from Cambridge, United States Contact this Seller
Item Unit Price Qty Price
RN2483 breakout (bare board) $3.00 2 $6.00
SHIPPING: United States Postal Service International First Class (Tracked) $4.00
Gnz Questions or problems?
The seller, Gnz, will ship your order from Zürich, Switzerland Contact this Seller
Item Unit Price Qty Price
IMST iC880a LoRaWAN backplane (Kit) $5.00 2 $10.00
SHIPPING: Swiss Post MiniPac International ECONOMY $9.50
Final order totals
shows an nRF51 and a Semtech SX1272
That's an interesting little device! And it would run Espruino :)
Doesn't say which nRF51 it is - there may be very little memory available for code - made even more difficult by the way the LoRaWan protocol would need implementing in it too.
I wonder if they're releasing an nRF52-based one - that could be a really useful module.
Yes: but on the other hand, there is a guy in Belgium that may be joining the Espruino for tangental interests and allow for some cool things.
Let's see if he is hooked.
He is capable of making a module with tangental interests in Britain: 100 POC units could be easy with the best intersection of interests and it is just a question of time
This might also be able to run Espruino. It's ESP32-based.
PyCom LoPy 1.0 - LoRa + WiFi + BLE
In TTN (and non-TTN), many use a Pi.
I would like to use a micro controller.
Having stated this in a TTN channel, someone asked:
"Are they (red: nRF52) powerful enough to run a packet forwarder? I'd rather use an ESP8266"
Two things here could be the hardware and the software.
From the point of hardware, and comparing apples with oranges, is there a golden bullet that says one contra the other?