TTN Gateway range

End Devices (Nodes)
1

I was under the impression that Lorawan gateways have range of 2 km in urban areas. However the gateways on TTN only seem to work if I am within few hundred meters range

I am wondering if it is because the gateways on TTN are of low/hobby quality and hence have limited range. If so, can someone recommend a good gateway that can deliver 2 km in an urban environment?

It would probably be helpful if the description for the gateways had relevant info about the gateway and its range in the TTN community map.

2

Not even a little bit, not slightly, not at all, just NO with a big NO on top.

All the chipsets in the world for a LoRaWAN gateway come from one place - Semtech. The technical work required to implement such a gateway is too hard for there to be a “cheap” version as such. The antenna, how it’s placed, the user interface, support, gold plating and marketing make up the rest of the costs.

For device radios, all the chipsets come from one place - Semtech. The technical work required to implement such a radio is too hard for there to be a “cheap” version as such. The antenna, how it’s placed, the user interface, support, gold plating and marketing make up the rest of the costs.

It is possible to choose cheaper / less optimal support components, but that’s easily filtered in the market place.

People do tend to read the marketing top number, never the bottom number - some super high quality gold plated Schwarzkopf encrusted gateways only manage a few hundred meters. Because it’s all about the obstacles between the device & gateway and their respective antenna. I live on higher ground than some of the rest of my town on the west side - so devices on the east side of town, more than 2km away, can get an uplink through just fine. But the ones in the town centre in the dip can’t.

So a gateway at the top of a tower block with a ropey antenna in an urban environment may well service a device 8km away - but none at the bottom of the tower block due to the way that RF radiates - you may want to do a little reading on polar diagrams which operate in 3D.

And it’s all about line of sight - if the device doesn’t have an unimpeded view of the gateway antenna and there is nothing big & sticky up in view of that path (read up on Fresnel zone), it should be OK with various caveats due to other sublties.

No one can do that without an RF survey - something we can comment on only if you give us your approximate location in this urban environment, where the gateway may be and where the device may be. And, assuming Google StreetView is available, it would still be advanced guesswork. And then the recommendation may just be a TTIG (cheapest gateway available) placed in a window. As all gateways are born pure, good is in their nature, it’s all about the nurture.

Ish, not really, due to all the above plus with it being a community gateway, it could have a poorly installed &/or placed antenna. What is of more use is TTN Mapper that does mapping - which is NOT range because if you walk around with a device doing mapping, you may be 3.14km with a really good RSSI & SNR and then, when you go behind a electricity substation, get nothing at all which happens to be big enough to block all devices downstream until it moves out of the Freshnel zone - so there can be a big blank spot for a gateway.

7 Likes
3

Thank you, Nick.

I had assumed that “urban” accounted for any obstacles and interference in cities and hence limited the range to 2km instead of the ideal 8 km.

Are there any case studies of a city wide lorawan network to understand how many gateways were needed to cover a certain geographic area? May be I could use that as starting point for research.

4

No such thing - it is what it is - the numbers are a very broad generalisation. My distance record is +118km using iffy antennas. There are a load of posts on here with people getting a few hundred meters.

Not really, probably only a very misguided attempt by academics who haven’t got out much.

Reality is everything - we do an RF survey for terrain, look at the bigger sticky up structures, come up with an educated guess, deploy some gateways, map it all, move some of them, map again, put in some in-fill gateways where there are bigger blank spots, rinse, repeat, watch a new building mess it all up, have new client put a tonne of devices where coverage is ropey, have a really useful gateway lose it’s home, see a new device deployment tip a gateway in to a level of overwhelm, have a new mobile tower introduce just enough interference to mess up a patch of coverage, new municipal refuse vehicles that start parking close to antennas, blocking signals for an hour at a time, etc etc etc.

All of this tends to happen in slow motion so it’s not a fire-fight.

Your very best bet is to learn about RF - antenna polars, Fresnel zone, signal strength vs distance (square root sort of stuff) - that’s a few hours well spent so you understand what is going on.

Then get a packet of Smarties, eat them, remove both end caps, cover the tube in Duck Tape, et voila, you now have a Line of Sight survey tube. If you stand where the gateway antenna is and look through the tube, if you can see a device location, you’ve a fighting chance of getting a signal through as long as there is nothing big and sticky up in that line, if there is anything big and sticky dangling down, it’s probably an alien mothership.

Then pick an area, deploy a gateway, map. Then move the gateway to somewhere topographically different, map. Do that again.

This should give you far more knowledge & understanding than any random “how to cover a city” type text. And with the aid of Google Earth / StreetView you can look at what the shapes are & which buildings may get in the way to extrapolate your learning.

Bear in mind that some buildings are relatively RF see-through and some buildings are unexpectedly RF blockers - it’s all in the construction - steel frame, concrete, metallic cladding etc.

You could look at TTN Mapper but you need to find an urban area that has a modest number of gateways - so Amsterdam (the home of TTN) doesn’t count. You want an area that has been reasonably well mapped and, using Google, look at the coverage vs terrain/buildings.

3 Likes
5

That would be impractical. Do you use the worst case scenario where all structures are RF shielding and there is a high level of interference? Or best case where all structures are almost RF transparent and there is almost no interference?

In practice a city will be a mix of both, depending on the size of the city and your location in it the balance will shift from one extreme to the other.

In modern buildings people struggle to get a mobile connection due to shielding. And mobile towers use a lot of power when compared to LoRaWAN. Your mobiles battery will last a few days max, LoRaWAN devices are designed to run for years using one battery so a lot less RF power is available and as a result such a building will severely limit its range.

1 Like
6

In most cases Gatreways and nodes are already set to use the maximum legally allowed transmit power.

The higher the antenna is located, and being sure its free of obstructions, then the longer the range you will get.

So called ‘urban’ areas can vary greatly, is it a flat ‘urban’ area of single storey bungalows or hilly ‘urban’ area with lots of skyscrapers.

By far the easiest way to learn about the practicalities of TTN\LoRaWAN propagation is to use a couple of simple point to point LoRa nodes. Locate a battery powered test transmitter in the typical place where the Gateway might be and wander around (cycling is good) with a portable receiver that beeps and displays RSSI\SNR when you receive a packet. You can check what happens when you hide behind a building, down a subway, in a wood, go to a first floor window, visit a hill on the outside of a city etc.

1 Like
7

@descartes @kersing @LoRaTracker

Thanks guys. That was very helpful.
I think I have a better understanding now and may be I was expecting a bit too much. I will run some more tests.

Do outdoor gateways perform better (in terms of range) than indoor gateways? Could the limited range I experienced be because they were indoor gateways and had too much interference right off the bat, as a result.

8

An outdoor Gateway with the antenna up real high and clear of obstructions will likely have a far greater range than an indoor Gateway in the ground floor of an office block.

2 Likes
9

Generally yes they do. Because the building will act as an RF shield. The reduction in signal strength will depend on the construction and materials used. In modern office buildings the coated (thermal insulation) windows are especially bad for RF penetration.

Think less interference, more RF shielding. Take a look at table 2 in this document to get a feel for the impact. (Keep in mind dB is a logarithmic scale)

2 Likes
10

Whilst the answers above are true, the science/physics answer is again a hard no. Same chipsets, same fundamental design, just different ‘clothing’ and the opportunity to get in a tangle with antennas.

An outdoor gateway is an indoor gateway is a weather proof case and usually has an external antenna giving you more choice, choice that requires you to understand antenna gain (so you don’t exceed legal limits) and cable losses (so you don’t run 10m of cheap coax from the ground floor to the roof and end up with half the power output you expect).

Conversely an indoor gateway is an outdoor gateway in a cheaper case, many of which have an antenna socket on the back. So by placing it in the right place with a short run of quality cable and a decent antenna mounted in a sensible place, you get an “outdoor” gateway without paying for all the bizarre concepts that manufacturers introduce with outdoor gateways - like metal cases that are IP67 when good polycarbonate IP55 under the eaves is fine.

In short, getting an outdoor gateway is likely to improve range & is a convenient form factor for many installs, but you can still mess it up and end up with poor coverage. So you will benefit from learning the fundamentals which I would strongly recommend you do as your very next thing so you can ask more informed questions - it’s OK to travel hopeful with yes/no questions like “outdoor vs indoor” but the devil is in the detail.

This video will cover much of what we are referring to:

https://www.thethingsnetwork.org/docs/lorawan/

For mental health reasons, don’t watch it all in one go. Have access to the pause button. A frosty or grape juice beverage will make it less of a fire hose.

1 Like
11

Increasing the spreading factor would increase the signal to noise ratio and very likely also the range. If you’re not already using SF12, give it a try.

1 Like
12

Give it a try, but it’s not a solution.

Partly because of the impact on air time which increases the chance of interference, the Fair Use Policy, battery life and the no small matter that the LoRa Alliance requires members to restrict the habitual use of SF 11 & 12.

1 Like
13

@kersing @descartes @LoRaTracker @stefankr

Thank you guys. I think you have given me enough to set realistic expectations and experiment.