Tragedy of the commons

Shared spectrum is not infinite, and license-exempt is not rule-free. This is the story of what a quiet listening node on 868 MHz revealed about the state of the mesh — and why a band without identity or coordination always drifts toward the same sad ending. Economists have a name for it. Radio amateurs have a cure for it.

One Sunday afternoon I configured an 868 MHz node to do exactly one thing: listen.

No beacons. No telemetry. No forwarding. Just a passive logger, quietly writing down every MeshCore packet it could hear.

The first surprise came fast: nodes more than 100 km away were landing in the log.

That part was delightful. LoRa's link budget is genuinely impressive, and 868 MHz carries a long way when a node sits high and in the clear.

The second surprise was less delightful.

A remarkable number of nodes were clearly not respecting the 10% airtime limit. Some were transmitting so often that the log looked less like a mesh network and more like a monologue with an audience.

And honestly? No surprise at all.

Spend ten minutes on the community pages and Reddit threads and you will find people openly celebrating the duty-cycle override. "Just switch off the 10% limit, your messages get through faster."

Faster. For you. For now.

There is a name for what happens next, and people have been writing about it since 1833.

The tragedy of the commons, in plain radio terms

In 1833, the economist William Forster Lloyd described a shared village pasture.

Every herdsman may graze his cattle on it. Adding one more cow brings that herdsman a full, private benefit. The cost — a slightly more depleted pasture — is shared by the whole village.

So each herdsman, acting perfectly rationally, adds another cow.

And another.

Until the pasture is mud and everyone's cattle starve.

In 1968 the ecologist Garrett Hardin gave this pattern its famous name: the tragedy of the commons. A shared resource, open to all, gets consumed by individually rational behaviour until it is useless to everyone.

Now swap the vocabulary.

The 868 MHz band is the pasture.

Airtime is the grass.

Every packet is a cow.

Each node that transmits more gets a private, immediate benefit: its messages get through. The cost — collisions, a rising noise floor, endless retries — is smeared out over everyone in radio range.

So the individually "smart" move is to transmit more.

And when everyone is individually smart, the band dies collectively stupid.

Why it always plays out this way

The depressing part of the tragedy of the commons is not that it can happen. It is that, without structure, it is the default outcome. The incentives are simply built that way:

• The benefit is private and immediate; the cost is shared and delayed. Flip the override and your message arrives right now. The degraded band arrives months later, for everybody.
• Anonymity removes the brakes. On a license-exempt band there is no callsign, no identity, no reputation to lose. Nobody knows it was you, so social pressure has nothing to grab onto.
• RF pollution is invisible. Your app cheerfully reports "message sent". It does not show you the packets of other users — or of the alarm systems and utility meters sharing the neighbourhood — that got trampled in the process.
• The negligibility illusion. "My one node hardly matters." Correct. Now multiply that reasoning by five hundred nodes in the same region.
• Success breeds congestion. The better the mesh works, the more people join, the more packets fly, the more collisions occur, the more retries pile up. Growth itself pushes the band toward the wall — we did the math on that in the 10% airtime article above.
• Feedback arrives too late. By the time the network is visibly unusable, the bad habits are already normalised, documented in forum posts, and baked into a thousand config files.

This is not a new movie. It played on the CB band in the 1970s, when 27 MHz turned into a shouting match. It plays on crowded Wi-Fi channels in every apartment block. It is currently in pre-production on 868 MHz.

Same script. Different frequency.

Three ways this ends badly

1. The network infarct: the hobby strangles itself

When hundreds of users in a region switch on the duty-cycle override, regional airtime does not creep from 10% to 12%. It jumps toward 50% and beyond.

The consequences are immediate and mechanical:

• Constant collisions. Channels get so congested that nodes continuously talk over each other. Packets are lost, resent, and lost again.
• Unreliability. Messages simply stop arriving. The network becomes useless for the one thing it was built for: dependable off-grid communication.
• The bitter punchline. By disabling the limit to "save" your own traffic, you helped finish off the network for your entire region.

A mesh does not die with a bang. It dies with ten thousand retries.

2. The regulator wakes up

Right now, regulators such as the BIPT in Belgium or the RDI in the Netherlands police the 868 MHz band mostly on the basis of complaints. As long as nobody complains, no measuring van leaves the garage.

Structural abuse changes that arithmetic quickly.

• Finding you is trivial. A node that transmits far above the limit is, by definition, a beacon. Direction finding is what these people do for a living. A spectrum analyzer picks a continuous offender out of the band in minutes, not months.
• The legal fine print bites. License-exempt use is only valid while the technical conditions are met — duty cycle very much included. Exceed them and you are, legally speaking, no longer operating a license-exempt device. You are transmitting without authorisation.
• Seizure and fines are on the menu. Inspectors have the authority to confiscate your lovingly assembled hardware on the spot and to hand out administrative fines that cost considerably more than doing it right would have.

The regulator does not judge your intentions. He checks the rules. We wrote a whole article about that mindset — it applies here with interest.

3. The legislator steps in

The 868 MHz neighbourhood is not a hobbyist ghetto. It carries real services: smart utility meters (in Flanders, Fluvius gas and water meters chat with the digital electricity meter over 868 MHz wireless M-Bus), social alarm systems, building automation, sensor networks.

If a license-exempt band gets structurally abused and those systems start misbehaving, the response is not a friendly forum post. It is political:

• Tighter rules. CE and RED enforcement can be sharpened, with pressure on manufacturers to lock firmware down and to bake the duty cycle into silicon — hardcoded, out of reach of any user setting, forever.
• Import crackdowns. Customs can start actively filtering uncertified boards at the border, raising the entry price of the hobby for everyone — including the thousands of people who configured their nodes correctly.

That is the cruel symmetry of the commons: the punishment for the freeloaders lands on the stewards too.

The culture shift: good configuration is a duty, not an option

"Everyone does it" has never once worked as a defence in radio law. Not on CB, not on freeband, and it will not work on 868 MHz.

The experienced corner of the community is now pushing hard for a culture shift, and the message is refreshingly simple: a correctly configured node is not a nice-to-have. It is the price of admission.

Every one of those settings costs you nothing meaningful and buys the whole region reliability.

Noticing that the mesh in your area is getting slow, or that packet loss is creeping up? Before blaming the neighbours: check your own node's airtime statistics first. Stewardship starts at your own antenna.

A commons can be governed — ask any ham

Hardin's grim story has a lesser-known sequel.

The economist Elinor Ostrom won the 2009 Nobel Prize for showing that the tragedy is not a law of nature. Communities all over the world have managed shared pastures, fisheries and irrigation systems successfully for centuries.

Their recipe is always the same: known members, agreed rules, monitoring, and real consequences for cheaters.

Does that sound familiar?

It should. It is a description of the amateur radio service.

• Identity: every transmission carries a callsign. There is no anonymous defection; your name is literally on the air.
• Agreed rules: band plans, coordinated by the IARU, tell everyone where which activity belongs.
• Monitoring: hams listen. A misbehaving station gets noticed, identified, and contacted — usually politely, occasionally with cake.
• Consequences: the licence that grants your privileges can be taken away. Skin in the game, as designed.

That is why a mesh experiment on amateur allocations can be coordinated: frequencies agreed, node density planned, duty discipline enforced by culture instead of by firmware. We can self-regulate, because the structure for self-regulation already exists and has worked for a hundred years.

On a public ISM band, none of that machinery exists. Anyone can key up. No identity, no accountability, no coordination — and "anyone" includes people with distinctly less noble intentions than you. You cannot coordinate an anonymous crowd, and you cannot phone a node that refuses to say who it is.

To be fair: the amateur route is a different game with its own rules — identification, no commercial traffic, encryption restrictions where they apply. But it is a game with a referee, played on a field the players themselves maintain. That is precisely what a healthy commons looks like.

Closing thought

The 868 MHz band is a wonderful shared garden, and Meshtastic and MeshCore are some of the most exciting things growing in it.

But a commons only survives its own popularity when its users act like stewards instead of herdsmen.

The tragedy of the commons is not fate. It is simply what happens when nobody takes responsibility — and it is entirely avoided when everybody takes a little.

So configure your node as if the band belongs to everyone.

Because it does.

Interested in more technical content? Subscribe to our updates for in-depth RF articles and lab notes.

Questions or experiences to share? Feel free to contact RF.Guru for practical RF and antenna support.