Choosing a network design is often what separates a good system integrator from a great one. Point-to-multipoint and mesh can both carry video and data from many sites, but they behave very differently when something goes wrong, and when the network has to grow.
The two designs at a glance
| Point-to-multipoint | Wireless mesh | |
|---|---|---|
| Paths to each site | One, through the central hub | Two or more, through neighbouring nodes |
| If a link or the hub fails | Sites behind it are cut off until it is repaired | Traffic moves to another path automatically |
| Line of sight | Every site needs a clear path to the hub | Each site needs a path to any neighbour, and relay nodes can route around obstacles |
| Interference | Wide sector antennas at the hub pick up and spread more interference | Directional antennas and individually chosen channels on each link |
| Growth | Limited by the hub’s capacity and fixed channels | Add nodes where coverage is needed |
| Best suited to | Small hub-and-spoke sites where some downtime is acceptable | Networks where downtime matters or that will keep growing |
How point-to-multipoint works
A point-to-multipoint network connects several remote sites to one central location, usually through a sector antenna that covers a wide arc. It is quick to plan and works well for a handful of sites with a clear view of one tall building or mast. The weakness is that everything depends on that central site and on each individual link to it.
How a mesh works
In a mesh, each node can reach more than one neighbour, so every site has more than one route back to the control room or core network. If a link fails or its quality drops, traffic takes another route. In an Anywhere network this recovery is sub-second and happens without anyone having to intervene.
What happens when something fails
With point-to-multipoint, a failed radio at a remote site takes that site offline, and a failed hub takes every site offline at once. Protecting the hub usually means duplicating its equipment. A mesh or ring removes the single point of failure because the alternative paths are part of the design.
Line of sight
Every remote site in a point-to-multipoint design must see the hub. Trees, new buildings and terrain often get in the way, and the usual fix is a taller mast or a separate link. In a mesh, a site only needs a radio path to a neighbouring node, and a relay node can be added to route around an obstacle.
Interference
A wide sector antenna listens across a broad area, so it hears more of the other radios nearby, and every remote site shares the hub’s channel. Mesh links typically use directional antennas, and each link can run on its own carefully chosen channel, which keeps interference down as the network grows.
Capacity as the network grows
All the sites on a point-to-multipoint sector share that sector’s capacity, so adding cameras at the edge reduces what is left for everyone. Conventional single-radio mesh has its own limit, because each hop shares airtime and throughput falls with every relay. Dual-radio Anywhere nodes avoid most of that loss and support more than 30 hops of backhaul, with multiple drop-off points to spread traffic across gateways.
Cost
Point-to-multipoint can need less hardware for a small site. As the network grows, the cost of duplicating the hub, adding masts for line of sight, and fixing outages tends to rise. With Anywhere nodes, there are no licence fees, so cost scales with the hardware rather than with the size of the network.
You do not have to choose only one
Most real networks mix designs. A ring or mesh can form the resilient backbone, with point-to-point or point-to-multipoint links reaching single cameras at the edge. Anywhere nodes support point-to-point, point-to-multipoint, daisy chain, ring and mesh on the same platform. Cabanatuan City in the Philippines, for example, runs ring, point-to-point and mesh links together to connect 300 CCTV cameras, flood systems, traffic lights and paging systems.
Questions to ask before you choose
- How long can each site be offline before it matters?
- Can every site see one central point, or will some need a relay?
- How many cameras or devices will be added over the next few years?
- How busy is the radio spectrum at the site?
- Who will fix a failed link, and how quickly can they reach it?