From Vancouver BC to Tijuana, a 1,500-mile chain of Meshtastic and MeshCore nodes almost connects the Pacific Coast — but California is the broken link. A comparative look at what it would take to close it, and why it matters.
Zoom out far enough on any Meshtastic node map and something interesting comes into focus along the Pacific coast. There are clusters — bright, dense ones — in Vancouver, Seattle, Portland, the San Francisco Bay Area, and Los Angeles. String them together and you've got the skeleton of something remarkable: a continuous mesh corridor running from Canada to the Mexican border. The spine of the American West, triangulated by cheap LoRa radios mounted on rooftops and hilltops by people who think decentralized communication matters.
Except it isn't continuous. Not even close. The clusters are islands. The gaps between them — especially through California — are hundreds of miles of uncoordinated silence. The West Coast mesh almost exists, and California is the broken link.
This is not a setup guide in the usual sense. There are no commands to run, no firmware to flash. This is a strategic picture — a survey of what's been built, what's missing, and what a realistic path to closing the gap looks like across three competing protocols. It's intended to be useful to operators in any of these communities who've wondered the same thing: can we reach each other?
This guide builds on the earlier Node Star field guides. If you're new to Meshtastic, Reticulum, or LoRa hardware, start with Vol. 1 (Build Your First RNode) and work forward. The analysis here assumes familiarity with basic mesh concepts and MQTT integration.
Let's start with the good news, because it's genuinely impressive. The Pacific Northwest has demonstrated that regional-scale mesh networking is not a fantasy. Seattle's MeshCore community — widely considered the largest MeshCore deployment in the world — has built a structured repeater network capable of routing messages across twelve hops. Operators have documented reliable paths connecting Vancouver BC to Eugene, Oregon: a span of roughly 350 miles, accomplished through deliberate infrastructure placement and disciplined configuration.
That's not a theoretical benchmark. That's people actually talking across state and national borders over LoRa radio, without cell towers, without the internet as a crutch. It works because those communities made specific choices: fixed repeater infrastructure on high ground, MeshCore's structured routing instead of flood broadcasting, and tight coordination between operators who treat the mesh as shared infrastructure rather than personal toys.
The Pacific Northwest proved it: a regional mesh isn't a fantasy. It's an engineering and coordination problem — one the rest of the West Coast hasn't solved yet.
Portland sits in this working corridor. So does Eugene, at the southern terminus of what the Pacific Northwest community has built. Below Eugene, the mesh thins out dramatically through southern Oregon and the long agricultural stretch of Northern California. There are individual nodes — there are always individual nodes — but no coordinated community, no shared infrastructure, no maintained backbone. The corridor ends at the Oregon border and doesn't resume until the Bay Area.
California has two of the most active Meshtastic communities in the country, and they've built impressive things independently.
In the Bay Area, bayme.sh coordinates a mesh of well over 150 nodes spanning the peninsula, East Bay, and surrounding hills — with infrastructure nodes on high ground including a pair on Mount Diablo's north peak. The community runs its own Meshview server, a Discord logger, and has coordinated a migration away from the default LongFast preset toward MediumSlow and MediumFast to reduce congestion and improve reliability.
Down south, SoCal Mesh has built a network of hundreds of nodes across the LA basin, operating as an area-wide off-grid text messaging system. Their stated philosophy is explicit: no cell towers, no internet — messages hop node to node.
Both communities are serious, growing, and well-organized. They just don't talk to each other — because LoRa radio has a range ceiling, and the gap between San Francisco and Los Angeles is about 440 miles past that ceiling. No antenna gain fixes this. No firmware update bridges that terrain.
The SF↔LA gap isn't a routing algorithm problem. It's a 450-mile geography problem combined with a coordination problem.
What could bridge it is an internet backhaul — a deliberate MQTT link between gateway nodes in both cities, operating on a shared private channel that both communities agree to monitor. This isn't a novel idea; Meshtastic's MQTT integration exists precisely for this situation. Both Bay Area and SoCal nodes with MQTT uplink configured are technically sending traffic to the same public broker at mqtt.meshtastic.org. They're just not on the same channel, and neither community has organized a shared private channel for cross-region messaging.
The Meshtastic project has implemented a zero-hop policy on the public MQTT broker — messages downlinked from the internet reach directly-connected gateway nodes but don't propagate further into the local LoRa mesh. The Bay Area community goes further, actively recommending the "Ignore MQTT" setting for most nodes and discouraging downlink altogether.
Internet feeds can flood the limited bandwidth of a LoRa mesh with far more traffic than the radio links can handle, particularly with poorly configured gateway nodes. The public server's zero-hop policy exists for exactly this reason. A private broker with a shared channel agreement sidesteps this entirely — private brokers aren't subject to the same restrictions — but requires coordination that doesn't currently exist between the two communities.
What doesn't exist anywhere in publicly documented California mesh activity is a deliberate, coordinated MQTT bridge between NorCal and SoCal with a shared private channel, agreed-upon settings, and active operators on both ends. That would require someone to build it.
MeshCore deserves an honest evaluation here, not a dismissal. Its structured repeater routing — where only dedicated infrastructure nodes relay traffic, rather than every device on the network — produces measurably less congestion and faster, more reliable message delivery in dense deployments. The Pacific Northwest success story is partly a MeshCore story. If you're building new urban mesh infrastructure from scratch, MeshCore's architecture has real advantages.
But for the specific problem of connecting SF and LA, the community gap is substantial. As of late 2025, Meshtastic's Discord had approximately 40,000 members globally versus around 3,500 for MeshCore — and MeshCore's North American momentum is concentrated in the Pacific Northwest. There are no organized MeshCore communities in the Bay Area or Southern California comparable to what bayme.sh and SoCal Mesh have built.
Choosing MeshCore for a California bridge means building a community and solving the bridging problem simultaneously, starting from near zero, while the existing Meshtastic communities watch from the sidelines. That's a much heavier lift than working with what already exists.
The more honest framing: Meshtastic and MeshCore solve different problems at different scales. Meshtastic is better for joining an existing metro mesh and contributing to community coverage; MeshCore is better for building reliable regional backbone infrastructure where you're standing up the repeaters yourself. A mature West Coast mesh probably needs both — MeshCore-style repeater discipline within cities, and a coordination layer between them that either protocol can serve.
There is no radio technology in the LoRa class that spans the distance from San Francisco to Los Angeles without a relay chain that would take years of infrastructure buildout. Acknowledge that constraint and the conversation gets cleaner: inter-city bridging requires internet backhaul. The question is which tool handles that backhaul most cleanly.
A gateway node in SF configured with a shared private channel uplinks to a self-hosted broker; a gateway node in LA subscribes and downlinks to the local mesh. Straightforward in theory. The key ingredients aren't hardware — they're a shared channel agreement between bayme.sh and SoCal Mesh operators, a private broker that sidesteps the zero-hop policy, and two committed operators who stay engaged on both ends.
Reticulum — the network stack underlying LXMF, NomadNet, and Sideband — was designed from the ground up for exactly this kind of heterogeneous, multi-transport networking. A transport node running on a Raspberry Pi Zero 2W with a Heltec V4 flashed with RNode firmware can simultaneously participate in the local LoRa mesh and bridge to a remote node over TCP/IP. Reticulum handles addressing, encryption, and routing across transport boundaries natively — no external broker, no zero-hop restrictions, no flood policy tradeoffs. Full multi-hop propagation in both directions, with encryption on every link.
If you're already running a Pi Zero 2W with RNode firmware — a common Node Star reader setup from Vol. 1 and Vol. 2 — you're most of the way there. Add a TCP/IP interface entry in reticulum.config pointing to a partner transport node in the other city, and you have a native bridge that doesn't depend on any third-party broker or corporate server. The network heals around failures automatically. This is precisely the use case Reticulum was built for.
The honest limitation: both paths require operators who are committed and present. The hardware and software are mature. What's missing is the human coordination layer.
A West Coast corridor that stops at San Diego stops 20 miles short of where millions more people live. Tijuana's metro population exceeds two million. Cross-border Meshtastic activity is documented and growing, with hobbyists on both sides already experimenting with nodes that reach across the fence.
The regulatory picture is worth naming honestly. Mexico's frequency allocation for LoRa devices technically differs from the US — 865–867 MHz versus the US 902–928 MHz ISM band — though in practice many operators on the Mexican side run 915 MHz hardware without incident. Radio signals don't stop at borders. A cross-border mesh is a real possibility, but it's also the most coordination-intensive part of the chain. It requires operators in Tijuana or Baja who are actively engaged, not just nodes that happen to sit near the border.
If building cross-border infrastructure deliberately, consult current Mexican telecom regulations (IFT) for the 915 MHz band. The hobby community largely operates in a gray area here. Ham radio operators with reciprocal licensing agreements have a cleaner legal path for cross-border experimentation and may be the right community to lead this segment.
A realistic West Coast mesh corridor doesn't require a single heroic infrastructure project. It requires a sequence of smaller, solvable steps — each with its own character and difficulty level.
The most radio-bridgeable segment below the working PNW corridor. Sacramento and the Central Valley have existing node activity. Solar-powered repeaters on high ground plus MQTT or Reticulum bridges for the longest spans could close this within a year with coordinated effort.
Requires a pure internet bridge — no continuous radio path through the Sierra Nevada and Central Valley is realistic. The key ingredient isn't hardware; it's a shared channel agreement and two committed gateway operators, one in each city.
The shortest gap and the most tractable. San Diego's mesh community is growing and geographically close to Tijuana. More a community connection problem than an infrastructure problem — a handful of operators on both sides of the border could close this quickly.
California has experienced catastrophic wildfire seasons that disrupted cellular and internet infrastructure across large regions. The Pacific Northwest sits on the Cascadia Subduction Zone, a source of anticipated megaquake activity that could disable conventional communications at regional scale. A mesh network that doesn't depend on cell towers or internet backbone isn't a luxury — it's a resilience layer that currently doesn't exist for the 50 million people who live along this coastline.
The Pacific Northwest community has already proven the model works at regional scale. The question for the rest of the West Coast is whether the communities that have built strong local meshes are willing to do the coordination work required to connect them.
The hardware exists. The communities exist. The firmware is mature. What's missing is the decision to build the bridge — and two people willing to hold either end of it.
If you're running nodes in any of the gap zones — Southern Oregon, Northern California, the Central Valley, San Diego — or if you're connected to the bayme.sh or SoCal Mesh communities and interested in inter-city coordination, Node Star is tracking the development of this corridor. Get in touch.
| Factor | Meshtastic + MQTT | MeshCore | Reticulum |
|---|---|---|---|
| CA community today | STRONG bayme.sh, SoCal Mesh | MINIMAL No organized CA groups | TECHNICAL Builders / hobbyists |
| PNW community today | MODERATE | STRONG Seattle is #1 globally | GROWING |
| Internet bridge support | Built-in MQTT — works well on private broker; discouraged on public server | Limited native support; community relies on radio relay chains | NATIVE Multi-transport by design |
| Intra-city reliability | Significantly improved since v2.6 next-hop routing | EXCELLENT Structured repeater routing | EXCELLENT Cryptographic addressing |
| SF↔LA bridge difficulty | MEDIUM Need channel agreement + private broker | HARD Build community first, then bridge | MEDIUM Need operators at both ends |
| Cross-border (Tijuana) | Possible; frequency mismatch is a consideration | Possible; same frequency considerations apply | Cleanest — TCP/IP bridge is frequency-agnostic |
| Best suited for | Joining existing metro mesh; community growth | Building new urban backbone with fixed infrastructure | Hybrid radio + internet routing across arbitrary distances |
| Resource | URL |
|---|---|
| Bay Area Meshtastic Community | bayme.sh |
| SoCal Mesh | socalmesh.org |
| MeshCore | meshcore.co.uk |
| Reticulum Network Stack | reticulum.network |
| Meshtastic MQTT Docs | meshtastic.org/docs/software/integrations/mqtt |
| Meshmap (node visualization) | meshmap.net |
| Node Star | nodestar.net |
The West Coast corridor isn't a hobbyist exercise. It's a resilience layer for 50 million people living in one of the most disaster-prone regions of North America. The Pacific Northwest community has shown it can be done. The tools are ready. The communities exist on both ends of the gap. All that's missing is the coordination — and someone willing to make the first call.