A meshnet (or mesh network) is a network where devices connect directly to each other instead of routing all traffic through a single central hub like a traditional router. Each device, called a node, can relay data to nearby nodes, which pass it along until it reaches its destination. This “hop by hop” approach means there’s no single point of failure, and the network can grow, shrink, or reroute around problems on its own.
How Data Moves Through a Meshnet
In a conventional network, every device talks to one router. If that router goes down, everyone loses connectivity. A meshnet flips this model: each node acts as both an endpoint and a relay. When you send data, it doesn’t travel a single fixed path. Instead, nodes cooperate to find a route to the destination, passing the data from one node to the next until it arrives.
The number of these intermediate steps is called the hop count. Fewer hops generally mean lower latency and faster delivery, because each hop adds a small amount of processing time. In a grid-shaped mesh with M rows and N columns, a packet traveling from one corner to the opposite corner needs (M−1) + (N−1) hops in the worst case. So a 10×10 mesh would require up to 18 hops for that longest path. Routing decisions constantly aim to minimize hop count, though sometimes a path with more hops over faster links can outperform a shorter path over slower ones.
As a mesh grows, the chance of congestion on any one link increases with more hops, while the overall network becomes more resilient because there are more alternative paths available. This trade-off between latency and redundancy is central to how meshnets are designed.
Routing: How Nodes Find Each Other
Nodes need a way to discover paths to every other node. Mesh networks use two broad strategies for this.
Proactive (table-driven) protocols continuously map routes to every reachable node, whether or not anyone is currently sending data there. The advantage is speed: when a node needs to send a packet, the route is already known, so communication starts with minimal delay. OLSR (Optimized Link State Routing) is one well-known example, and it’s been tested extensively in real-world setups using small, inexpensive computers like Raspberry Pis as mesh nodes.
Reactive (on-demand) protocols only figure out a route when someone actually needs to send data. This saves bandwidth because the network isn’t constantly exchanging routing information, but there’s a brief delay while the route is discovered. The IEEE 802.11s standard for Wi-Fi mesh uses a version of this: a node broadcasts a route request, neighboring nodes forward it, and the destination sends a reply back along the discovered path.
Some protocols blend both approaches. Babel, defined by the Internet Engineering Task Force, is primarily proactive but includes reactive features. In testing, it performed roughly twice as fast as another popular protocol called BATMAN, partly because of its efficient path selection. BATMAN, for its part, works by having nodes regularly broadcast small status messages. That constant chatter can congest the network if links are very reliable (paradoxically, it scales better when some packet loss naturally limits the flood of status updates).
Meshnets in Your Home
If you’ve bought a multi-unit Wi-Fi system for your house, you’re already using a mesh network. Products from companies like Google, Eero, and TP-Link place multiple access points around your home, and they relay traffic between each other to eliminate dead zones. Your phone or laptop connects to whichever node has the strongest signal, and the nodes handle the behind-the-scenes routing.
Smart home devices take this further. Zigbee and Thread are two wireless protocols specifically designed for low-power gadgets like light bulbs, sensors, and door locks. Both use mesh topologies, and both can support hundreds or even thousands of devices on a single network. Every smart bulb or sensor that stays powered on can act as a relay for its neighbors, extending the network’s reach without any extra hardware. Thread has become particularly prominent as the backbone for the Matter smart home standard.
VPN-Based Meshnets
The term “meshnet” has also been adopted by VPN providers. NordVPN’s Meshnet feature, for example, lets you link your personal devices (or friends’ devices) into a private, encrypted network that functions like a local area network, even if the devices are scattered across different cities. You can route your internet traffic through a friend’s device so it appears to come from their IP address, essentially turning any device in the mesh into a mini VPN server. This is useful for accessing region-locked content, playing LAN games remotely, or securely sharing files without uploading them to a cloud service.
Security in a Decentralized Network
Because data in a mesh passes through intermediate nodes you may not control, encryption is essential. The standard approach, as outlined by NIST, is for any two hosts that want to communicate to first establish a secure tunnel between them (using protocols like IPsec). This means that even if a relay node in the middle is compromised, it can only see encrypted data it can’t read.
Commercial mesh routers handle this transparently. Your traffic between nodes is encrypted by default, and you manage the network through a single app. In community or DIY meshnets, the burden falls on the users or administrators to ensure encryption is properly configured at every node.
Where Meshnets Are Used
Home Wi-Fi coverage is the most familiar use case, but meshnets show up in far more demanding environments. Emergency responders deploy portable mesh nodes when existing infrastructure is damaged by natural disasters, since each node only needs to reach one neighbor to join the network. Military communications rely on mesh topologies because they keep working even when individual nodes are destroyed. Cities have experimented with municipal mesh networks to provide public internet access without running cables to every block.
Community meshnets, sometimes called “community wireless networks,” are built by volunteers who mount antennas on rooftops and link them into a neighborhood-wide mesh. These projects aim to provide internet access independent of commercial ISPs, and they’ve gained traction in areas where traditional broadband is expensive or unreliable.
Industrial settings use mesh networks for sensor monitoring in factories, mines, and agricultural fields, where running wires is impractical and devices need to cover large areas with minimal infrastructure. The self-healing nature of a mesh (if one sensor fails, traffic reroutes through others) makes it well-suited for environments where maintenance access is limited.
Strengths and Limitations
- Self-healing: If a node drops out, traffic automatically reroutes through remaining nodes. No single failure takes down the network.
- Easy to expand: Adding coverage is as simple as adding another node. No new cabling or reconfiguration of a central hub is required.
- No single bottleneck: Traffic spreads across multiple paths instead of funneling through one router.
The downsides are real, though. Each hop adds latency, so in a large mesh, communication between distant nodes can feel sluggish compared to a direct wired connection. Bandwidth also decreases with each hop, because relay nodes share their capacity between their own traffic and the traffic they’re forwarding. And while mesh networks are easy to expand in theory, performance can degrade as more nodes compete for the same wireless channels. In dense deployments, careful channel planning becomes necessary to avoid interference.
For most home users, a mesh system with three to five nodes handles these trade-offs well, covering a typical house with minimal latency impact. For large-scale deployments, the choice of routing protocol and careful placement of nodes become critical to maintaining performance.