Improving my Gateway range and reach by changing antenna

Hello All,
I am using a The Things Gateway and The Things Node to test some range and transmission reach in urban area, in Morocco.
Let me give you a quick visibility on my installation:

  • My Gateway is plugged via ethernet to my wifi router and is next to it (same table, <1m away)
  • My Gateway is on the balcony (so outdoor), at a 4m altitude from the ground (1st floor basically)
  • All the testing has been done with The Things Gateway and The Things Node as they come straight out of the box.

Considering this, I have achieved a range of around 500m radius. I would like to increase that range. To do so, I am considering replacing the Things Gateway antenna by a stronger one, do you think it is a good idea ? if yes, what antenna should I highly consider ?
If this is not possible, is there any gateway on the market (e.g Kona Mega Ex IoT Gateway from Tektelic or others) that have way better range or antenna performance ?

To give you further understanding of the situation, I would consider 2.5km range as great in urban area (this is an urban area with mostly villas and not many buildings above 9m height)

Lastly, can the lack of range associated to the Things Node and its internal antenna performance instead of the Gateway situation ? What is the average range performance you get with the Things Node in an urban area ?

Thank you very much !!

imho the gateway used does not have much influence on range you can cover. Much more important is the position of the antenna. The antenna should be on the roof (or as high as possible). The antenna should “see” the nodes because the propagation at 868 MHz is like light with some reflections from buildings.
I would not recommend an antenna with to much gain, 3-5 dBi is sufficient. The position and the radiation-pattern of the antenna is more important than the gain.
I am also in an urban area and use an indoor gateway (Dragino LPS8) with an external antenna on the roof connected by 10m coaxial cable (Aircell 5). 2.5km is no problem if the node has an appropriate antenna (1/2 lambda) and is 2m above ground.
Additional hint: There are some other very interesting topics in this forum dealing with antennas.

Hello Wolfgang,
Thank you very much for your answer ! Could you further develop on the antenna you used.
I am not very familiar with the “antenna” field, did you make the antenna yourself ?
What do you mean by 10m coaxial cable ? 10 meters ? If yes, this length seems huge to me, how do you install that, towards where does the antenna point, and what is an appropriate positioning ?

I will check other posts around the subject in the forum, meanwhile, I had people mentioning to me that going just from the things gateway to a The Things Outdoor Gateway or something from the Tektelic brand such as the Macro or the Mega would improve the range dramatically, what do you think ?

Going from an indoor antenna to an (properly placed) outdoor one will improve range enormously. The gateway hardware is all based on the same reference design and is far less of a factor in the equation.

There’s a whole antenna topic here:

That’s confusing - you said you were connected to your router on Ethernet but the only other Things Gateway other than the outdoor one is the indoor one that doesn’t do Ethernet. Which actual gateway do you have?

You’ve overlooked the original Kickstarter gateway which has both WiFi and Ethernet. :grinning:


Maybe the KS gateways are on the 2nd user market in warmer climes now!

Depends on where you are, in most places in the World you cannot simply replace the current antenna with a ‘stronger’ one, if you want the gateway to remain legal.

In most places your limited by regulation to 14dBm Effective Radiated Power (ERP) and its assumed your using the minimum practical antenna which is a 1/4 wave vertical with radials or a dipole. These antennas have a ‘gain’ of 2.1dBi.

If you use a 5dBi antenna directly on the gateway, so minimal connector or cable losses, you would need to reduce the Gateway transmit power by 3dBm to stay legal (Gateways are normally set to max power, 14dBm). Although the 5dBi antenna will receive nodes at greater distance because of the required reduction in Gateway transmit power, the gateway will not be able to contact those greater distance nodes and TTN is now not working as intended.

In effect there is little point in adding a ‘stronger’ antenna to a Gateway to get better performance, unless it is your intention to operate the gateway outside legal limits.

The exception to using a gain antenna, is that it is permitted to compensate for connector and cable losses, but then if you have those you wont be receiving more distant nodes in the first place.


Hello ! As mentioned below by kersing I have this gateway:
I bought it new, weeks ago, maybe the new gateways are way better, let me know :wink:
I am fine with buying a different one if needed.
Thank you for the threads, I will go read them.

Firstly, thank you very much for your explanation, however my question was more is there a better antenna than the one we receive with this gateway ( that remains legal while improving the range of this gateway.
If not, can you share any gateway that have tremendous results in terms of range, and is supposed to be outdoor.
My nodes will tend to have STM32WL chips to transmit and internal antennas (if this information can help you understand my situation) and I am trying to achieve 2-2.5Km range in suburban/urban area

Thank you Jac ! When you say “properly placed” you mean on the roof, with no obstacles surrounding the antenna, or is it something way more complex and detailed ?
Also if I take my gateway and put it on my roof using an extension cord and put some of this antennas instead of the given one ( am I supposed to see tremendous increase in range ?
Finally what do you mean by “improve range enormously” ? times 2 ? times 5 ? even more ?
I know you have only a little of information and I am asking multiple questions, thank you for your help, any answer is welcome !
Thank you

There is not a ‘better’ antenna that will improve performance and stay legal, since you need to reduce the Gateway transmit power.

All Gateways basically use the same Semtech LoRa technology and will have the same range capability.

What makes a major difference to gateway range is the placement of the antenna, as high as possible and as clear of other objects\buildings as possible.

And whilst I see you asking questions about how much range will improve by better antenna positioning, thats difficult to be precice as so much depends on where the antenna currently is, where its to be moved too and the environement you are in.

Once using the exact same equipment (@434Mhz) I measured a range of 1km for a signal in a flat urban area and > 1200km when the same (transmitter) equipment was in Earth orbit, such is the vast difference there can be in range depending on the environment.

I did once carry out some experiments on node placement in a flat area twhere I live that was a mixture of woodland and housing. The difference in signal strength between a node at shoulder height and on top of a 10m pole was around 10dBm, which equates to around 3 times the distance.


As you mention the antenna should be placed with no obstacles between it and the horizon. Any feeding antenna cable should have appropriate cable loss. Appropriate is where the cable and connectors reduce the signal to the extend that the resulting combination equals the original antenna properties (a generic dipole).

The improvement in range depends on the environment. Are you able to mount the antenna with a unobstructed view of the horizon or are there taller objects (buildings, trees) shielding the signal in one or more directions? Will nodes have a line of sight to the antenna or does the signal (still) need to pass through (multiple) buildings? Is there any structure reflecting signals?
RF propagation in an urban area is hard to predict.
I’ve had deployments where placing the gateway antenna outside increased the range from <2km to >10km but in other cases the difference was from about 300m to 500m.


Indeed so, very variable indeed.

1 Like

Surely you do not know much about antennas, but you must bear in mind that it is a balloon inflated with air and the same content of the balloon you are going to distribute.
So if you use a monopole it is as if you put the balloon and all the energy is distributed to the sides (something like a donut), and if you use a half lambda or quarter lamnda it will be like the same donut full of spikes.
For this reason what I understand you are trying to achieve is that the part of the donut is as flat as possible, for this and from experience I would advise you to use a 1/2 lamnda and what you should keep in mind is to put a good ground system reference, so you can put a metal plate or join a good grounded chassis, since if you do not do that the shape of your radiation will tend to air upwards and not to the sides, with what energy is lost pointing towards the sky and therefore you do not reach more than 500 meters. In the same way it must be in a high place.
I live in Ecuador, South America, here we work in the 900 MHz band. But with the antenna that I tell you (lamnda 1/2) I have achieved distances of more than 2 km without having to increase the amplification or violate any transmission norm from my country.

There is not a ‘better’ antenna that will improve performance and stay legal, since you need to reduce the Gateway transmit power.

Sorry but I don’t think this is the case.

I think only half of the equation has been analyzed here… Gateway -> Station isn’t the most common transmission either; of course a better antenna, be it higher gain, or not using omnis (resort to sectors instead) will definitely have an impact, same goes for placing it higher or on a better strategic position.

As long as your gateway sets the gain of the antena in order to compute legal PIR, there isn’t any drawbacks, but actually quite the opposite:

  • Gateway radio will be able to “hear” better remote stations, much more limited in transmission abilities, without resorting to high amplification gain (higher noise…) and this is what matters! bear in mind 90% of traffic is station -> gateway
  • Reducing required radio Tx Power allows it to operate better, less distortions, making gateway -> stations transmissions more reliable.

So definitely, antenna gain and quality is #1 factor on a Gateway effective range.

I don’t agree with you. Radiation-pattern and position of the antenna is more important than the gain. I do not want to break distance records. I want to supply a stable signal to the nodes and receive signals from the nodes around me.
If the antenna of a gateway is on a (high) roof, the high-gain antenna beams over the nodes on the ground but it receives the signals of nodes 100km away. This is the reason why some antennas have a “down-tilt”.
imho there is no “best” antenna. The antenna should fulfill the individual requirements concerning radiation-pattern and gain.

1 Like

And how will those limted/remote nodes (note not ‘stations’ in LoRaWAN parlance) hear the join accept from the OTTA join request - the effective TX from the GW is constrained in law and nodes will not have high gain antenna assist hence will not hear the join accept even if the GW can hear the faint join request…just one example of where you will fall down. Also remember we are discussing TTN - a community network, hence any advantage you gain for yourself should not be at the cost of causing problems for other users. Another point is higher gain antennas start to move even further away from the classic, desirable isotropic radiation pattern, away from the classic donut even, and worse start to introduce nulls of lower gain notches where you get coverage shadows in closer range to GW even as you reach further out to the horizon with the main beam (GIYF). Please take note of the sage advice from other Forumites and stay at lower gains - there is a reason why most emergency services use simple ant’s (e.g. half dipole/dipole ants with 2-4dBi gain to ensure even coverage in the regions of concern/interest, other than in specialist circumstances. Also be cautious as badly configured higher gain ants can have unintended up or down angles - fine if talking to distant aircraft or mounting on a high (broadcast) mast or on hills and mountains, giving coverage below, but also then limiting horizon reach.


I disagree.

First off, by definition, if you make an omni directional antenna ‘better’ than a simple 1\4 wave vertical or 1\2 wave dipole, you are increasing the antennas ERP, hence the need to reduce TX power to stay legal.

Placing the antenna in a better\higher location can improve effectivness significantly, but ERP is not impacted, so you dont need to reduce TX power to stay legal.

And finally, if you uses a better Gateway antenna, and reduce the TX power as a requirement to stay legal, then indeed the better antenna will receive nodes from further distances, but the reduced Gateway power means that those more distant nodes cannot receive messages from the Gateway.

1 Like

Do you have any evidence to substantiate that ?