The Arboretum

Mesh Networking

4 min read

Mesh Networking

A network topology where every node connects directly to its neighbors rather than routing through a central hub — and one of the few infrastructure models that maps naturally onto how communities actually work.

The Idea

Most internet infrastructure is radial: a provider owns the backbone, leases the last mile, and every household terminates at that provider’s edge. The topology is hierarchical and brittle — when the center fails, or raises rates, or decides a neighborhood isn’t worth serving, the periphery has no recourse. Mesh networks invert this. Nodes connect peer-to-peer, traffic finds its own path through the web, and the network can route around failures without any central authority to petition.

The structural argument for mesh isn’t speed or cost, though both can improve. It’s that the topology matches the social reality. Neighborhoods have houses next to houses. Apartments have line of sight to other apartments. The physical proximity that makes communities communities is also, it turns out, the substrate for a network you can own together.

Origins

The technical concept predates community applications by decades — distributed packet routing was a design principle of ARPANET, chosen deliberately for resilience. But the community mesh idea took shape in the early 2000s, when cheap 802.11 hardware made it possible for households to buy their own equipment and route it without ISP permission.

Guifi.net in Catalonia is the largest realized example. Started in 2004 in rural areas that commercial providers wouldn’t serve, it grew to over 37,000 active nodes by 2021, operating under a Wireless Commons License that encodes participants’ rights and obligations: you can use the network, and you must extend it to others who need it. The Foundation registered as a licensed telecommunications operator in 2009, giving the commons legal standing.

NYC Mesh operates a volunteer network across New York’s boroughs, peering at internet exchange points rather than buying commercial transit, with members mounting routers on rooftops and connecting to neighbors through line-of-sight links. Detroit’s Equitable Internet Initiative, run by the Detroit Community Technology Project, trains Digital Stewards — community members who learn both wireless engineering and community organizing — and built networks reaching hundreds of households in Southwest Detroit, Islandview, and the North End. The pairing of technical and organizing training is not accidental. Networks need maintenance; maintenance needs people who have reasons to show up.

Althea Networks developed a compensation model where nodes pay each other micropayments for bandwidth in real time, attempting to solve the sustainability problem through market mechanisms rather than voluntarism alone.

The Tension

The maintenance problem is real and has killed more community mesh projects than any technical failure. Volunteers age out. The person who understands the routing table graduates or moves. The access point on the roof of the community center fails during a rainstorm and no one knows whose job it is to climb up there. Guifi.net solved this partly through scale — with 37,000 nodes, there are enough people who know what they’re doing to keep the fabric alive. Detroit’s Digital Steward model tried to solve it by building the repair capacity into the community from the start, not importing outside experts. Neither approach is a full answer.

The compensation model Althea proposes introduces a different risk: turning neighbors’ connectivity into a market. The rooftop node that was a gift to the block becomes a revenue stream, and the person who can’t pay gets deprioritized. Whether market mechanisms sustain networks better than solidarity does is still an open empirical question.

Connections

  • local-first — the principle that infrastructure you own locally is more resilient
  • appropriate-technology — mesh hardware as convivial tool: repairable, understandable, locally owned
  • commons — the Wireless Commons License as a commons governance instrument
  • mutual-aid — community mesh as infrastructure mutual aid
  • calm-technology — low-intervention networked systems that don’t demand constant attention

Further Reading

Graph View