What's your judgement on LoraWAN for non cellular IoT networks

Living in Amsterdam, Netherlands and having backed & eagerly awaiting the Electron which ultimately relies on large telco’s, I came across LoraWan as an alternative data network: The Things Network in my city recently set up such a network at low cost, free to use and owned by the community. Check out their wiki to learn about boards supporting LoRa. Also note that The Things Network initiated a KickStarter campaign to fasttrack cheap availability of both gateways and nodes. A node does what the Electron does but hooks up through LoraWan. I am backing a node, which I hope to have by mid 2016 - and will then compare the Electron telco + dataplan approach with this one.

My question to the Particle community: how do you judge LoraWAN for non cellular IoT networks?
Apart from Smart Cities and taking things in your own hands: seems ideal for Remote Data Sensing in rural areas without 2G/3G/4G.

Thanks, Repke

4 Likes

Repke, I think LoRaWan is going to be right way to publish IOT data to the cloud. I am really excited about the possibilities. We don't have much coverage here in USA, but SigFox is in several cities and several other providers are saying they will be rolling out next year.

seems ideal for Remote Data Sensing in rural areas without 2G/3G/4G.

I think it is ideal for urban areas also, the G networks are overkill for such small messages.

LoRa is coverage in Belgium is now in all major cities and by the end of 2016 the hole country will have coverage (Evolueer naar een slimme onderneming | Proximus). So yeas, it is a way to go for a lot of devices to be connected. Does particle have any plans in the direction of a LoRa module?

I don’t know, I sure hope so. Maybe someone from Particle will answer.

LoRa in Italy lives in Milan, and stops there for now.
You know, we are a Banana Republic for a lot of things :angry:

In the end LoRa promises to be more pervasive than cellular networks, at least until the 868 MHz band will remain uncrowded for a while.

The LoPy currently on Kickstarter should help spreading coverage further. I would love to see a Lora device from Particle!

I’ve done some reading about LoraWAN. I have some Lora radios (FCC).
What I’d like to know is: What are people paying, going to pay, for network access. The network operators are for-profit.

LoRaWAN is a very interesting technology, but I am curious to see how such a wide area network will work in an unlicensed band. WiFi works in an unlicensed band as well, but it is short range. If multiple vendors are setting up competing networks in the same band that cover large areas that overlap, it will get crowded. Not to mention all the other devices that work in the same band (cordless phones, baby monitors, etc).

Now, LoRaWAN is extremely low bandwidth and designed to deal with interference, so it will certainly be ok up to a large capacity in an area. Many, many sensors would take up the same bandwidth as one WiFi camera streaming video. But as more and more sensors come online and perhaps demand more and more bandwidth, things start to contend.

It is sort of similar to cellular networks, they started out with the same topologies many years ago. Now, as spectrum becomes more scarce and more and more devices demand more bandwidth, resources are running low. So the industry has tried to adopt technologies that supplement long range deployments with short range networks. This is seen with technologies such as HetNet and LTE-U.

Just the cycle of things, I guess. Start with wide-scale deployments until resources are used up, then attempt to move back to smaller scale as the large scale networks reach a critical mass. LoRaWAN would certainly work well for many years as-is, I’ll be curious to watch and see if it cycles back.

Ed Hightower has made several public presentations comparing the various low-power wide-area-network technologies, including LoraWAN. There’s video of one from last month on Vimeo. Several of these networks are going live in the USA this year, making for interesting times.

If you prefer to read about it check out his blog at IoTandBeyond.com, which has linkes to slides from his presentations.

Disclosure: Ed is a friend of mine.

– Pete

1 Like

The 968 MHz SRD band that LoRaWAN uses in Europe actually has limitations on the duty cycle of the device (1% or 10% depending exactly which channel used) so the regulators have thought ahead a bit. If you employ some kind of dynamic frequency selection and power control you are not limited by the duty cycle requirement but I suspect the first generation implementations will not use these features. An interesting question is if the duty cycle requirement is enforced by the different vendors' firmwares. If not, there will be mayhem...

I am also hoping for more crowd-sourced networks like in Amsterdam (I know there are other cities as well) and cheaper hardware. But this is coming, see e.g. LoPy mentioned previously. In the crowd-sourced model, however, the difficult part will be to get a reliable network server and backbone running.

Lora/LoraWAN would be a perfect fit for Particle!

Yeah, a Particle Boson with a LoRa chip onboard would be fantastic.

I’m also thinking about getting a LoRa gateway to cover our town. A community owned free IoT network like the Things Network is a fantastic idea! Let’s take back our communication infrastructure!

2 Likes

Sorry to revive an old thread, but it’s just to clarify (and switch off the excitement a bit):

In Europe LoraWan regulation on the 868 MHz band imposes an 1% duty cicle, so you must listen for 99% of the time, and can transmit only for 14 minutes in a round of 24 hours.

See TTN fair access policy

3 Likes