Zigbee network intermittent issues

Question

My Zigbee network connection seems to be intermittent. Suddenly all routers will lose connection, for example unable to ping any of the routers. After awhile (e.g overnight or a period of a few days) the network will be up again.

Set up:

• 1x coordinator
• 21x routers
• 42x end devices(in use)

Currently using channel 25.

I want to find out if mounting the Zigbee router on metal partitions(wall) will cause problems for the network. The part of the network that is in another building with no metal partitions seems to be more stable. Or are there any other factors that will cause the network to be unstable?

Answer 1

It is important to identify whether the problem is :

• Devices or network deployment
• Devices or network configuration.
• Devices Hardware (or firmware versions)
• Environment
• Security configuration

The paths to take here are too diverse so you need to narrow down the cases. Sometimes a photo helps to spot the possible deployment issues.

Here are some ideas:

1. A nearby lift can reduce the signal due to the amount of metal in it
2. Also consider the water as strong signal attenuation, this includes humans, to if that is a crowded area it is better to have the antennas above 2m.
3. I have seen communication problems and it was due to the wiring was badly bended in a sharp 90 degree angle, this damaged the internal filaments.
4. Another case I saw is that too many Zigbee gateways (Access Point) can produce conflict since the clients may roam from one to another, this depends on the local distribution of the gateways and the signal strength. If the clients are statically located I would assign them fixed gateways
5. The proportion between the amount of clients and the amount of Gateways has to me calculated and check.
6. Don't forget to check also the configuration, maybe there is a bottle neck, this is, many clients use the same subnet, so all the packages are forwarded to a single switch/router. In that case splitting the clients in groups or subnets and diverging the traffic can be a possible solution.
7. Are all the client devices and gateways running the same firmware (SW) versions? Are all up-to-date
8. Is there any difference in the security configuration? The encryption and VPN usually slow down the traffic if we compare them with non-protected networks.
9. I almost forgot to say that the logs from the clients or the Gateway will provide you relevant information about the communication. If all the net goes off abruptly and remains in that state for days, I will go for the gateway logs an see if the is some indications that the communication is lost.
10. Check for duplicated IP address, the behavior you describe can also be due to 2 devices with the same IP and when one of them is on the net, the other is not communicating at all. Check the routers IPs in this case.
11. Once I experienced communication lost in a customer and it turned out that the security of the net did not allow clients to reconnect more than 3 times in a row, then the router would ban my client for hours. All this was security configuration in the router/gateway.

But of course for all this we need more information about the deployment and the configuration