Two topologies, two philosophies
In point-to-multipoint there is a central station, such as an antenna or an access point, and each node communicates directly with it. In a mesh, each node can communicate with several neighbors and forward the traffic of others, so that a message reaches its destination through successive hops.
Practical comparison
| Aspect | Point-to-multipoint | Mesh |
|---|---|---|
| Coverage | Each node must see the central station | Extends through hops between nodes |
| Obstacles | A node without line of sight is left isolated | Traffic goes around the obstacle |
| Redundancy | If the central station fails, the whole network goes down | Alternative routes exist if a node fails |
| Latency | Low and predictable, a single hop | Grows with the number of hops |
| Capacity | The central station’s capacity is shared | Each hop consumes channel time |
| Complexity | Lower | Higher: routing, testing, diagnostics |
Why mesh is not free
Forwarding consumes the radio medium. With nodes sharing the same channel, a multi-hop message occupies the air several times, and useful capacity decreases with the number of hops. Forwarding nodes must also stay awake, which complicates designs with battery-powered nodes.
When each one makes sense
- Point-to-multipoint: when the nodes have line of sight to the central station, predictable latency and throughput are needed and simple management is preferred.
- Mesh: when the geometry is difficult, there are changing obstacles or tolerance to failures of individual nodes is required, and traffic per node is modest.
What to check first
Before choosing, it is advisable to carry out a coverage study at the site, know the traffic per node and the maximum acceptable latency, and define what happens if the central station or a key node fails. A poorly planned mesh network can have bottlenecks at the nodes close to the gateway to the rest of the facility.