hello, if i have a device outside gateway range can i make a device repeater? or a gateway repeater?
is there any documentation about that?
best regards
Luis
If you want to install a repeater, why not just install a gateway covering that area?
To install a gateway i need internet, correct? I am not able to use internet outside gateway range because i dont have it.
Correct, always better if you can add another gateway.
You can always build a Lora to Lora link and then link via serial to a node connected to TTN.
I am using ESP32. want to send GPS information from several GPS devices outside lora gateway.
Then i can send ESP32 using ESPNOW, or MESH to a ESP32 that can send to gateway.
What will be better? ESPNOW or MESH, is there anything written out there?
There are a number of mesh protocols in the RF24 library which I believe you can run on top of a LoRa radio so you can ask Google to find the details & options. There is no reason why the middle device canât reformat the payload and send it as a LoRaWAN packet. But you will need to embed the originating deviceâs info in the LoRaWAN payload.
As for a LoRaWAN repeater, youâll find extensive prior discussions about the feasibility of this - I believe the last detailed one was for someone who needed to get around a mountain for a couple of nodes - suffice to say âitâs complicatedâ and weâve not seen reports back of anyone succeeding. Whereas the first paragraph basically describes a scheme that does pretty much the same thing.
I have something like that running but using nRF24L01âs for the mesh and then the centre node sends as LoRaWAN as that reaches the gateway some distance away.
You should probably search up and read the existing threads which discuss the ways in which reporting GPS position over LoRaWAN is already a packet length / usage limit challenge at moderate to longer distances within a gateway coverage area.
Hi, thanks for the answer. i googled about RF24 looks amazing yes, didnât knew, Thank you.
My problem is, if i use RF24 to communicate between devices i only have ~800m range. I am trying to track cattle that can be far from each other⌠I think, that several ESP32 with LoRa (one on each animal) sending to a ESP32 LoRa receiver and this one catching the gateway will send the data to the cloud . Just need to register one device(receiver) and will send in the payload as variables the id of each remote device. What do you think of trying this?
So if this one node is accepting GPS location inputs from many other (local) devices, what are your plans to keep the one TTN connected node, acting as a relay, within the 30 second air time fair access limit ?
I just need to send the information once each half an hour. And if goes out of a geofence limit.
Montana or something, maybe, but last I saw, cattle are herd animals, being almost half a mile apart would be pretty unusual. And youâll be surprised what a good antenna high up can achieve.
Alternatively, having sown the seed of Googling, you could look at other libraries that do mesh protocols with other radios, including SX1276. There are at least two others.
Dual core 32 bit is going to need some battery. Last I saw, changing the batteries on a cow can get a little fraught. Something very low power would be better, perhaps with a solar top-up.
Nah, Iâm good thanks, mesh networks are a whole bag of excitement that I only do once a year.
But hereâs some tips from tracking teenagers on expeditions:
- Data format that records who was in a geographical radius as a cluster, where & at what time, each one holds the same record.
- Relay the data to any other nodes at every opportunity, but track who you hear & who youâve sent it to so you donât waste power sending if you donât need to.
- Figure out a plan for your GNSS power budget.
- Stick âgatewaysâ at obvious locations (not takeaways or shops with confectionary, that only works with teenagers, Iâm thinking water or food sources).
- Gateways positioned where they can relay between each other to the edge to get to the interweb are good.
- Gateways with backup GSM are good too.
- Set aside a stupid amount of time with half a dozen friends running around the country side for testing.
One other possibility I have to try out is to put solar powered mini-gateways in a sort of grid / on high ground / in dips, in range of their neighbours and then ditch GNSS and just use RSSI to get an approximation of where the tracked item is - after all, if youâve got a rough idea down to around 100m where they are, itâs a bit hard to hide a cow. Then the âtrackerâ on the animal just needs to transmit its ID (briefly) at appropriate intervals.
But this is mostly off-topic for TTN.
How can I register a 32u4 feather device from adafruit on the platform? It appears to me at the activity âneverâ. How would I solve the problem?
Wise to plan that the node sending the locations into TTN stays within the fair access limit at SF12, in which case you would be limited to a total of 21 transmissions per day.
If there were only 2 animals having packets relayed, that would be 96 packets per day.
I know someone who tried that, not that successful.
The farmer could tell when an animal had dissapeared (been nicked) but by then it was too late.
Same fundamental issue as a car/vehicleâŚonly comes into its own and has value once itâs been stolenâŚat which point you really do want the tracker to work! 
Unless manglement are using to track say delivery drivers or if drivers are taking excessive breaks etc. For farming community there is an extra value that some of my clients have noted which is they can track herding behaviour and monitor for over grazing etcâŚmoving animals on if it risks depleting yield or over grazing to extent it delays land recovery etc. So value goes beyond inherent black market price of a cow I guess 
Whereabouts? What sort of terrain?
Absolutely - which then enters the Twiglet Zone of Argument (TZA) about the Apparent Total Coverage of Gateways (ATCG) that anyone creating or deploying trackers thinks exists. So thatâs GSM as a fall back.
But if you have wide open spaces (like Kansas, but weâre not in Kansas anymore), and rounding up your livestock is an issue - like roughly where are they and where has the ânaughtyâ pack gone - then I think this has possibilities.
But as above, useless for cattle rustling. Now where are my leather chaps.
Is this a thread hijack? If so, do let us know and weâll move it to a separate topic.
Exmoor âŚ
1 Like
ActuallyâŚKentucky! (And I think they were also later targeting Oz & Nz) One of my early client LoRa apps was doing tracking/monitoring of horses on very long, but thin ranchesâŚ30-60 miles by 3-10miles? approxâŚfascinating task. I think solution in the end revolved around neighbouring ranches cooperation to cover most of an area approx 100 x 50mile area with a near 2 dozen GWâs/data collection points. A sample of animals was tracked & monitored initially as herding gave good indication of where >90% of them were at any time. IIRC they only applied Pareto to look for the approx 20% of wanderers at a later date, as these were brood mares brought into early season by shining uv light in their eyes using a simple hood to simulate onset of longer days*âŚtracking wasnât a problem as these horses were carrying big batteries to power the Leds so a gps fix a few times a day was no burden!
Another fascinating LoRa/LoRaWAN application example brought to you byâŚThe Twiglet ZoneâŚstay tuned for next weeks episode⌠
(*) a long story around the economics of race horses for over a beer when people finally get to meet
1 Like
I would dearly love to organise a large area with a number of adjoining farms each with a decent antenna mounted high up to provide good coverage - that would provide a brilliant testing ground. I found just one farm tends to need some infill gateways (particularly in hilly north Derbyshire), usually on LTE which then becomes another ongoing cost rather than a capital purchase.