Industrial gateway for rural areas with poor cellular coverage

I will be lucky if I too find your experiences or suggestions on the use of LoRa technology in rural areas with poor cellular network coverage.

which industrial LoRa IoT gateway is best suited to this case?
the distance between my LoRa sensors and an area covered by a cellular GSM network with or without 3G coverage (Internet) in which the gateway will be set up is estimated at 20 to 30 km.
Do you think I should set up several gateways to reach my destination?

the Kerling iStation gateway which offers an operator sim card, how does it work? I don’t need 3G internet?

What suggestions of LoRa components might I have from you if I want to test LoRa transmissions early without using Gateway?

please answer me
thank you

It’s not clear why you think one gateway would be better than another in a situation where you don’t have a backhaul to connect it to.

Concentrate first on how you intend to achieve the backhaul connectivity.

Wires? Fiber? Point-to-point microwave link?

Something like satellite Internet might be possible in theory, but may get unreasonable on a cost basis, and may introduce a latency which challenges current design assumption of TTN which require the backhaul network to move traffic in a small fraction of a second.

2 Likes

well noted
thank you very much for your reply.

yes indeed you are right, I have not thought of another wireless or wired data transmission technology to set up for the transmission of sensor data in rural areas, because I think they are more expensive and difficult to establish.

hence the choice of LoRa transmission.

yes, I asked for information on the existing IoT gateways to make sure that no LoRa IoT gateway could exceed a given coverage distance: like 15 km indicated by several manufacturers of LoRa IoT gateways.

So that’s why I thought about why not interconnect two to three gateways to have coverage that exceeds 15 km to join a backhaul? What do you think?

otherwise, what do you think of the following architecture sensors - to LoRa gateway - to BTS GSM - to 3G Internet?

excuse me, I am a beginner in this field

I am waiting for your answer

thank you in advance

It’s theoretically possible, and often referred to as a “repeater”.

However it is not practically possible with many current nodes, as it introduces a time delay their firmware often isn’t built to to handle.

Even if the node could, during a join, be commanded to use a larger receive window delay, nodes are joined to the network as a whole, not to an individual gateway. TTN uses a 1-second delay, and isn’t prepared to tell a particular node “well, you only have connectivity through a repeater, so you need a longer delay”

Right now, to use TTN, you need low-delay Internet access at the gateway locations. There are in theory non-TTN schemes which could perhaps work around that, but non-TTN schemes are not really on topic on the TTN forum.

thank you very much for your answer it cleared things up for me.

however I am a little discouraged that the LoRa transmission may not be the adaptable technology for my project which must be in very large rural areas. I must think of another solution. What do you think?

however, it corresponds to my IoT project in the city center.
hence my questions:

I ordered Things Uno which are with me now I must also order my Things gateway as soon as possible but I wanted to be sure of their operation in my country (Africa) which has not yet put regulations on the use of free frequency bands,

However I do think that I can communicate my Things Uno to the Things gateway using their own antennas for example with the 868 Mhz EU frequencies mentioned on the Things Uno. What do you think?

and for the Uno things i need an external antenna or is the internal antenna enough for me to connect it to the gateway?

I tried to register my Uno things on thethingsnetwork platform but it sends a join denied response, is this due to the absence of the network gateway or is it another problem?

also I am wondering if I am using a low cost Ultrasonic Prototyping sensor I will not find accurate results ?. better to order the real more expensive Ultrasonic level sensors for more efficient results. I want your advice please.

other question on the LoRaWAN network server of the ThingsNetwork at some point I must switch to a pay account to view my sensor data? I want to have some information on that.

thank you especially @cslorabox
and for all

That’s a bunch of very different questions.

Probably it would be helpful to spend some time understanding how a LoRaWAN network has three parts, and what each does: the node, the gateway, and the server.

On the other hand, it may well be a perfect solution. But as we only know you want to cover a reasonably large area that has little or no mobile coverage, we will struggle to advise you further.

  1. What are the sensor devices going to sense?
  2. How many of them?
  3. How often do they need to transmit?
  4. How quickly do you need the data back at the office?
  5. Where is the office? ie rural or some distance from the area.

It is possible to have LoRaWAN gateways standalone on solar power. These could then relay the data by other means at various speeds depending on your needs.

thank you for the answer and the hope you gave me.

quite simply, I want to monitor several zones where each zone regroups several pools, the information sought is the water level in each pool continuously.
in one area the ponds are not more than 2 km from each other.

however the gateway should be placed where there is 3G cellular network coverage.
whence the question?
the pools are very far from network coverage, I wanted to use repeaters either between the sensor network and the gateway or between the gateway and the center (office).

This is why I asked if there are IoT gateways that have a larger coverage (more than 15 km) or to suggest a component that serves as a relay.

I familiarized myself with LoRa components and I am very passionate about this technology that’s why I wanted to use it in this project.

Thank you for your answers

thank you for the answer and the hope you gave me.

quite simply, I want to monitor several zones where each zone regroups several pools, the information sought is the water level in each pool continuously.
in one area the ponds are not more than 2 km from each other.

however the gateway should be placed where there is 3G cellular network coverage.
whence the question?
the pools are very far from network coverage, I wanted to use repeaters either between the sensor network and the gateway or between the gateway and the center (office).

This is why I asked if there are IoT gateways that have a larger coverage (more than 15 km) or to suggest a component that serves as a relay.

I familiarized myself with LoRa components and I am very passionate about this technology that’s why I wanted to use it in this project.

Thank you for your answers @descartes

This is the flow:

[sensor]–àLora–>[Lora IOT Gateway]----->3g cellular/Wifi---->[LoraServer/TheThingsNetwork]

The sensor transmission of data to a Gateway is quite large and ideal for rural areas…however your question sounds like it is a celluar coverage issue of the gateway.

[quote]I ordered Things Uno which are with me now I must also order my Things gateway as soon as possible but I wanted to be sure of their operation in my country (Africa)

which has not yet put regulations on the use of free frequency bands,[/quote]

Which Country? Africa is a continent:)…ive seen a few post from people in Africa running Lora gateways with no issues.

as suggested in a few comments above i suggest spending more time studying the technology and architecture

Having worked with a gateway with satellite backhaul I can confirm it didn’t work for 90% of the downlink traffic due to timing issues. Avoid if possible, if not look for another LoRaWAN network as it won’t work on the TTN public network.

In your use case I would check the Lacuna Space offering. They use satellite LoRaWAN. That is one way (sensor to back end) only but should work if you only want to send sensor data to a central point. (That central point will require internet access to access the data)

1 Like

Really, why - stop assuming things and perhaps help us to help you by answering ALL of the questions. Gateways can act standalone, it may be there are other ways of getting the data to where it’s needed.

Coverage of gateways can be increased depending on the terrain and how high you can get the antenna of the gateway. All gateways have the same basic chipset, but some are better designed than others.

It may be that a design that uses a local LoRa Peer-to-Peer network that then relays data on to a gateway by using a directional antenna could be used. There are many possibilities if we have the info.

Not while participating in TTN they can’t. And this is the TTN forum, so when someone says a “gateway” it’s reasonable to assume they mean a TTN gateway.

Non-TTN schemes aren’t really on-topic.

But a

would be.

1 Like

Probably not:

  • The technical design of the peer-to-peer LoRa network isn’t a TTN subject, and is where most of the work would go

  • Re-injecting the entire local network’s traffic through TTN is likely to exceed the allowed airtime under the acceptable use policy, especially at the slow spreading factors likely needed.

Really, if one wanted to do this, the link back to something with Internet should also be a peer-to-peer link. And then, just as in the case of people who periodically ask about how to inject traffic from Ethernet based sensors, the answer is that such traffic should not be injected into TTN but rather flow into the user’s data backend by a parallel path.

It’s an interesting problem, indeed. But it’s not about TTN.

I’ve gone trough similar setup in a 10times smaller scale, but the problems seem to be the same as @HassinaMauritanie i.e. shitty gsm coverage and nodes in rolling terrain.

IMHO, 2-3 gateways with big antennas ~10dB the biggest antennas legally possible in you region/country with (separate) gsm backhauls and nodes with big antennas would make it possible. Well configured backhaul would work even on 2G/EDGE - know it from experience. 20-30 km in open terrain should not be a problem for SF12 packets, but it must be tested physically.
And whatever you do, 1 gateway is rather not enough. Having redundant gateways would decrease the probability of timing errors (apart from other advantages).

To alleviate the problems with GSM coverage consider a gateway with a socket for external GSM antenna or find a good GSM router and use any gateway with WiFi/Ethernet backhaul. The WiFi/Ethernet in the middle can introduce additional lag to the communication but it can be compensated by better GSM connection and more freedom in relative placing of the two big antennas: the LoRaWAN and the GSM.

As @cslorabox wrote, the LoRaWAN repeater is no practically possible, but if can program microcontrollers … you could develop a LoRa - LoRaWAN bridge-node. You sensors use plain peer-to-peer LoRa to send data to your bridge-node and the bridge-node is on the other side a regular LoRaWAN node and forwards the received data to the GW. In this scenario, the you would have to develop you own protocol between the sensors and the bridge, which can be nontrivial if it’s to be secure and reliable but on the other hand, very interesting.

And at the end, I will not mention other approaches, like Meshtastic or RadioShuttle, because they are not related to TTN :wink:

And whilst it might work, readers should take note that this would not be legal in most places, for a transmitter.

What wouldn’t be legal? 2-3 gateways or antenna +10dB?

A gain antenna of +10dBi for a transmitter.

2 Likes