I am starting with LoraWan/TTN and there’s something I don’t understand about TTN : say I want to communicate between 2 devices that are 10km far from each other (one at home and one at work). Device A is in a zone without surrounding gateways so I need to setup my own. Problem is how do I know that my gateway will be able to relay messages to surrounding gateways in order to reach Device B? I understand TTN is a community mesh LORAWAN network therefore there’s no guarantee 1 gateway will be close enough to another one in order to relay messages, is that right? Is there a way to ‘simulate’ a message travelling from point A to point B with all known TTN available gateways so that we can anticipate whether the message from A will actually reach B?
LoRaWAN isn’t a point to point system or mesh. Devices transmit, any gateways in range that hear it relay the uplink to the main servers that decrypt & pass it on to your server.
Gateways know very little - they are not routers.
You would best served reading the Learn section (link top-right of this page) to get a better understanding of what LoRaWAN is.
If you need to move a message from one place to another, due to technical issues of downlinks, this may not be the technology for you to use.
A gateway is not able to relay messages. It receives the signals of the nodes, demodulates them and if it is a valid LoRaWAN -signal forewards them to the TTN-server by an internet connection.
Buy a gateway, deposit your node and find out whether you (your gateway) can receive it. If your gateway can see the node, you have a good chance to receive it’s signal. The propagation is “quasi-optical”.
Use LoRa P2P - outside scope of this forum… TTN is based on LoRaWAN and as such devices do not communate directly with each other…
They don’t as above - GW’s communicate over the interent to a LNS they do not communicate with each other…
You need to follow the Learn section and get better understanding of LoRaWAN and how it is architected and works before planning your application and use case - it may be it is not the right technology for you…
Wrong again! See Learn (More a star or stars network no mesh involved…
rest of your question (s) therefore moot…
In fact I did read the Learn section before posting but what I completely missed (among other things ) is that GW are not communicating between them, they are ‘just’ forwarding messages received from the node to Network server.
What I get from your answers is that there’s no downstream possible to another Lora device/node.
In fact what I would like is to send data from a Lora sensor located in a place with no internet (or say a place where I cannot setup any internet connection myself).
It could be P2P (but challenging when the 2 devices are 10km from each other) or via an application through TTN but it would then be mandatory that my Lora sensor is able to have a direct link to a GW.
Yup, that’s pretty much every LoRaWAN device - long range / low power radio.
Yes, that’s how it works, device transmits and one or more gateways hear them - otherwise they’d be transmitting out in to the void and be pretty useless.
If we knew what information you were trying to get from one end to the other we may be able help better.
Does the ‘other end’ have internet?
As for the 10km, LoRa is good, but sometimes it can’t defeat local radio use by others or the landscape is not right (you need line of sight) - so even P2P may not work.
To be honest at this stage I am more experimenting than anything else but here would be the use case:
An end point device (let’s call it nodeA) based on TTGO board (Lora chip) with a temp/humidity sensor attached , self powered with 18650 battery. This end point does not have any internet connection (at least I cannot setup a gateway close to it and there’s no available gateway around)
The other end does have internet
10km between them (not LoS unfortunately)
At first when I heard about Lorawan I thought I could use community ‘gateways’ to hop and carry data from one device to another one. Now I understand this is not at all how it works.
In my use case instead of having another end device I could retrieve the temp/hum information from a cloud application (I guess that’s part of the principle of LoraWan: the node is sending to the LoraWan network and you can retrieve your data using a computer/Api/ etc… as it is exposed on an Application server )
This would be fine but as NodeA does not have internet then I guess there’s no real solution.
I had a look at Meshtastic project. It is very interesting but in my case I cannot setup a 10km mesh network between the 2 locations
I have probably been biased by YT videos stating Lora can communicate at several kilometers distance. It is probably true at LoS but in real life applications it is not always possible to have P2P with 2 devices at LoS. I watched some videos showing that in very busy environement it would drop to just a few hundred meters or even less.
The paradox for me (please excuse my lack of knowledge) is that in order to use lorawan you need internet…but if I have internet I would probably not need Lora.
STOP with this already. The really poor choice of a TTGO board which has hard to implement low power modes would use the LoRa radio to transmit which would be picked up by a gateway. It would not be on the internet. It happens to have a WiFi radio on board but that’s because it uses the ESP32 chip, not because a LoRa/LoRaWAN device needs to be on the internet.
NO LoRaWAN devices are normally attached to the internet. They exist because there is no internet nearby. They send out a radio message. Which is why we say that they transmit. Gateway picks it up, passes it to main server which then forwards it to where ever you need that, yes, does generally need internet.
The paradox here is how complicated it has ended up in your thinking because you are fixated on the sensor device needing internet. They do not.
If it turns out your device’s location is too far from an existing gateway, then you’ll need to arrange a gateway close enough that is on the internet. But even then the gateway could be using mobile internet.
We won’t be commenting on other solutions, as this is a TTI funded forum to support TTN. But some of your additional thinking isn’t on track either. And a fair amount of the YT videos are based on people with only entry level experience, so are supposition. Some of us can get our P2P going over 100km’s. LoS is pretty essential, but then there are other ways & means. But none of them involve the device having both a LoRa radio and an internet connection. Otherwise what would be the point?
If you are running a battery driven device, use something like an Adafruit Feather M0 with RFM95 which has a very small sleep current. Or be prepared to learn how to make a TTGO sleep with LoRaWAN. Or schedule regular battery changes. You can start out with the TTGO and see how you go with battery life. If you already have one, why not set it up and get going - use LMIC-node - and build on the experience.
Thanks again for your time.
I just realize the way I explained is very confusing (English is not my native language, sorry about that).
I clearly got the fact nodes are not connected to the internet.
What I meant to say is that the node needs to be communicating with a gateway (using Lora protocol) and this gateway needs to be connected to the internet.
In my case, either I use an existing gateway (which is not possible because there are no GW available around/in range with my sensor), or I setup my own GW…which needs to be connected to the internet. This is why I was mentioning this paradox: I am using Lora but I need internet nearby for the GW to operate. If it happens my sensor is using wifi like the TTGO then I would rather use its wifi to connect to the internet with some sort of mobile data router nearby instead of a Lora GW.
Of course I clearly understand my thinking is based on the fact that the node does have wifi connectivity which might not be the case for most lora nodes.
That’s a choice, just run a WiFi hotspot or cut out the WiFi element and have the sensor directly attached to a GSM/LTE modem.
One of the reasons to chose to put a LW gateway on mobile data is because it can then support a considerable number of small battery driven sensors. And with appropriate placement (high up), it could provide coverage over 2 to 3km radius = ~30km2.
The best way to figure all this out is to try things out. Good luck!
) is that GW are not communicating between them, they are ‘just’ forwarding messages received from the node to Network server.
