Using high-gain YAGI (11-23dBi) antennas for LoRa gateway (DON’T!)

The noise level, seen by the receiver, seems relativly constant in my experience as in the recever is seeing general background noise with probably widespread particular geographical sources. So in a city for instance where you point a directional antenna may make no difference to the level of noise seen.

Easy test to carry out, if you had access to the great outdoors, my guess is that there will be little to no difference.

In any case you would be implementing a Gateway with very restricted coverage.

Thank you guys for all your helpful comments! I will stay with a standard gateway but it’s important that questions like mine can be discussed here in a ‘liberal’ fashion :)…I am not intending to break any regulations, that’s why I was reaching out for your help!

Why is that important? And why more important than us spending time answering questions from other users? Because that is the choice you are making by requesting our response.

I wanted to say that I think it is important to have an open discussion without people attacking each other or with aggressions…

Sure it is.

However when posts are made that have ‘suggestions’ that are illegal or contrary to the LoRaWAN standard, then the issues will and should be highlighted.

I am surprised that some of the ‘suggestions’ have been allowed to remain, its easy for a casual reader to get the wrong idea as to what is not permitted in a casual read of threads such as this.

I changed the topic title to reflect the outcome. You should not use a directional antenna.


Generally, of course. Unless you actually want to receive packets from far away such as in the case of LoRa enabled life jackets or other critical devices used at sea.

Like this one of mine, proudly pointing out over the Irish Sea.

Understanding the use case before dismissing someone’s question and falling back on TFU docs really doesn’t help.

No-one has a network just to have a network, they have one because they want to do something with it.

Abrasive comments don’t help the community, they only serve to divide it and we work much better when we seek to understand the other viewpoints.

TTN is a community of tens of thousands and has grown because what of what it enables and supports.

Long Range forever.20200616_205735

1 Like

I assume your LoRa-enabled life jackets are not using LoRaWAN? Hence, this is not connected to TTN? All fine, of course!

That’s fine, unless that network is connected to the TTN Community Network, which is what this forum is about. On TTN, don’t do this please.

Understanding other ‘viewpoints’ is not and never was the issue.

Regardless of how compelling the use of gain antennas might appear, it needs to stay legal.

Of course it is legal. Why wouldn’t it be? Fully complaint ETSI SDR.

Saving lives at sea is pretty compelling especially when everyone else is pointing antennae inland and not at large expanses of water.

Perhaps we should develop SoRaWAN too?

This is not down town Groningen, it has low population density to say the least.

Being abrasive is very unhelpful.

Don’t do what Arjan?

How the community decides to use its own gateways whilst obviously complying with local regulations and TTN guidelines is down to them is it not? It’s all LoRaWAN btw.

Why send a HAB up to break the world record if not to celebrate the technology’s potential.

The LoRa life jacket concept is still in development but thank you for the keen interest. Getting a SFx message out from a device of a user in peril or a submerged container / obstruction to shipping is possibly more compelling than LoRa enabled mousetrap traffic.

My point is to apply some context and to try and avoid being abrupt and preachy. It seems like you’re also saying don’t put your gateways up high incase it becomes reachable.

SRDs is not what Nicolas had in mind when he authored LoRa or am I.missing something?

I am fully aware to which community this forum connects but thanks for the reminder.

Long live long range.

Saving lives at sea using LoRa technology is certainly a very worthy use case. However please keep in mind TTN community network is not the right platform for life critical services. If you search the forum you will find messages by TTN core team members providing that statement.

TTN community network has no SLA and over the last year suffered several outages lasting hours to days. Not a technology to use when lives depend on reliability.

That is right. And other people using the network (rightfully) assume a gateway that receives transmissions of their nodes will be able to send data to them as well. Creating an asymmetric setup voids that and may lead to a lot of hair pulling as to why things don’t work as expected. So while it may serve your use case, it might hurt someone else’s.


Sure, and as much as LoRa enabled lifejackets are nice application, a yagi pointing out to sea isn’t in the TTN guidelines unless you are trolling us and it’s actually on LoRa which would make so much more sense for side stepping join procedures & fair use policy. HAB (which I do) doesn’t use LoRaWAN as a norm.

What if one of my horse rider trackers was activated but couldn’t relay the position because it was tangled up with a gateway with a directional antenna? If we are going to take the moral high ground, I suspect my use case will be required far more than a mousetrap or a submerged container …

Trolling? Us?

Neither, I am merely replying to someone elses comment that you have chosen to subsequently respond to so how is that trolling.

ABP doesn’t have a join procedure does it? I use OTAA with persisted keys.

In terms of HAB, I was referring to the TTN world record attempt recently not yours, unless that was you? Very cool if so, still very cool if not.

I dont think you get my use case unless that is your horses swim out to sea. Most horse riders probably have phone or other connected device with GPS and reachable by bue lights.

There’s no moral high ground and I’m not sure who the collective we is or are that you refer to… If anything unmatched omnidirectional antennae cause much greater traffic overhead at the network level as they have zero coverage planning or propagation analysis prior to being turned on.

What I suggest merely focus EM at patches of sea, land and sky where there is no LoRaWAN coverage. All within ETSI guidelines and duty cycle. Just because a HG antennae is capable of breaching ETSI doesn’t mean that it will.

Whilst you’re out on a canter, you might want to consider that those that could benefit most from LoRaWAN don’t own a horse and live tens of kilometres if not a great deal further from the nearest gateway.

Frankly, the notion of message entanglement and horse riding is not something I suspect will affect anyone and I think the mousetrap wins that one mate.

I know Jac. I’m just merely pointing out that using directional antennae is super valid in some cases. I don’t propose a network full of them, but at the edges and to span distance. They offer exceptional opportunities.

Egalitarian networks are good enough when combined with an easy to use app like TTN Mapper for example.

It only takes one successful send to save a life or cover large areas full of IEDs and to pinpoint unexploded ordnance and a vast multitude of similar uses cases. I could go on and share experiences of how this technology really can save and enrich lives.

Their use should be considered where certain constraints mean there is no TTN coverage and where SLAs are luxuries people can live without.

Always happy to support healthy debate.


Please don’t use high-gain antennas on gateways connected to TTN. Because of:

Of course, the following is clear and sounds harmless:

I appreciate that your nice use case does not seem to hurt TTN, today. And I assume offshore wind turbines, oil rigs and container ships would never want to rely on the TTN Community network either. But can we really think of all future use cases? And how would I test such antenna setup; would it still point towards the sea while I’m doing my development? But worst of all: the next person will also think their use case is special and warrants some non-standard setup connected to TTN (or justifies abusing the shared radio spectrum), unknowingly hurting others.

Why not start your own network, using LoRaWAN or some bespoke LoRa-protocol, for use cases that would need a non-standard LoRaWAN gateway setup?

I don’t think we need guidelines for everything. For example, should TTN have a guideline saying that a gateway is not allowed to ignore downlinks for an unknown DevAddr? Or that gateways should be active 24/7 during normal operations? While such guidelines are not in place, I would not expect anyone to configure such downlink filter or opening hours.

I don’t know if the sending node was somehow not LoRaWAN compliant, hence might have affected other devices. (Which indeed would be bad.) But surely no gateways were adjusted for that experiment?

1 Like

Indeed, the use case would appear to be ideal for the establishing of a private network setup and one where the remote nodes were not required to recieve packets at all. Then there would be no problems, legal or otherwise, for using high gain antennas on the Gateways to increase range. Simple.

1 Like

Of course it does not harm TTN, to suggest otherwise is very unhelpful.

Providing TTN coverage where it does not exist such as at sea, remote areas and in developing nations is not harmless, it is essential if we are to to seek to democratise access to and for a wonderful community.

Innovators wishing to test and trial their ideas won’t have the option to always afford a private network so to suggest they stay off TTN really does go against what the community is all about. Please correct me if I have gotten this wrong.

Most organisations of the type you describe will want to run low volume trials to limit cost and prove the concept before making any significant investment. Proving a concept on TTN and then switching up to TTI or other private networks is a path we have already taken for duty cycle benefits whilst we still happily enable coverage for budding entrepreneurs by having TTI branded gateways offering TTN coverage.

Justify abuse of the radio spectrum? Unknowingly hurting others? I am not talking about single channel devices, but even those devices are valid in certain use cases but not on a set freq. My suggestion here is that some better traffic analysis might have helped identify the problem sooner and given that the end customer is a Council, why should TTN even be used, a basic SDR dongle would show whats happening locally? If anything, the link you quote appears to demonstrate a private company using TTN to test it’s field trials validity without checking for any private LoraWAN or RAW LoRa.

I’m afraid I just don’t accept your narrative relates to what I have suggested Arjan.

We are a TTI customer for private networks and have always sought to promote it’s development in recognition of the fantastic job TTN has done and continues to do.

As far as guidelines are concerned, we should certainly have them for key aspects that cause pain to other users. After all, the gateways are the property of the owners and spectrum management will soon become an issue unless enforced ADR or duty cycle is baked in. Even some basic signal data analysis can identify packets coming from a long way away (or through buildings, granted) and from devices such as the SF12 “mousetrap” that is spamming two of my TTN gateways, it would be smarter to allow the console user to ban a certain DEV-EUI / ADDR should it be shown to breach ETSI and / or TTN fair use.

The only way to protect an egalitarian network is to offer GW owners some control over the traffic it allows i.e. ban for 48 hours and then double that each time the cycle is abused. I agree that in areas of high node / population density that interference at scale is an issue, I am not proposing using YAGIs in locations where coverage exists.

My key point is that no-one has the answer to all possible use cases and that the use of CE certified radio equipment within published regulatory limits and acceptable use policy is entirely legitimate.

As far as the HAB is concerned, very very cool. Non-standard gateway set up and why would they need to be adjusted? There would have been a distributed local spectrum impact and a heck of a lot of de-dup happening but a great way to validate the technology.

If LoRaWAN can operate from space, then I am pretty sure that some strategically placed gateways with well tuned / matched directional antennae covering large swathes of low or no TTN coverage. would expand the richness and reach of what is a powerful tool and a great community.

Vive le network mon ami.

1 Like

So who decides, if implenting something is ‘not harmless’ it implies it causes harm to someone or something ?

If coverage does not exist and has never existed or existing signal data equals value X (where X indicates lat lon or other specified location parameter ) is possibly a decent start.

Also, a quick look at a map of TTN and known private LoRa coverage might be faster. Link budget and TX / RX are really simple noddy ways to check coverage and teach an ML instance to spot excessive dedup requests etc. So establishing simple LoRa zones is one way to ensure that anyone running HG that interefe today can be locked out by the network server. As more coverage comes online and as the local RF dynamics change so can the acceptable use policy.

It’s not for me to decide anything other than how to operate my gateways within legal and fair use policy guidelines.

The brainpower of the engineering faction in the community have probably already solved this using electrons not words ;-(