Looking for cnWave 60GHz real-world feedback

Hi all, just looking for some honest, straight up opinion on cnWave 60GHz. On some of the wisp forums I am seeing operators take down their cnwave and replacing it with Tachyon. One operator in particular commented on constant cn disconnects from dn’s that are only 100m away when light snow and rain come in. They swapped out the cnwave deployment for tachyon and all issues went away. I am also reading that 60ghz cnwave is a huge flop/disappointment. But with all things there is always 2 sides of every story.

Hello @DigitalMan2020 we are aware of the post and I would love to help them and understand the issues they are facing, it’s certainly not the norm.

We have thousands of radios deployed around the world in some extreme conditions, Canada (with lots of snow) has one of the largest cnWave deployments with over 4500 radios in one network alone.

Tachyon like UI Wave is an 802.11ad device, there are certainly differences between the air interfaces, but I would still stand by that cnWave offers far more for density, scalability and redundancy which is tougher for these .11ad devices to do beyond simple hub and spoke deployments.

1 Like

We’ve deployed a reasonable number of both cnWave and Wave radios, approximately 800 cnWave and 300 Wave radios. Our opinion is that the 2 are significantly different tools for different applications. Primarily due to the increased range of the upper channels and the 5 GHz backup, we’ve deployed Wave to upgrade speeds for current 5 GHz customers. We’ve been generally pleased with reliability out to about 2 miles. If the 5 GHz radio and failover worked better, we’d feel better about customers out to 2.5 mi or maybe a little further.
Our cnWave deployments are in manufactured home parks with very high densities. One example is 500 lots in .125 sq. mi. and we have 300 customers at peak in that property. I don’t believe any other wireless technology could handle that density. The Terragraph technology is designed to maximize frequency reuse and minimize self interference. However, as we’ve learned the hard way, the design work is not trivial. It doesn’t seem too complex on paper, but in the real world with obstructions, customer take rates, etc. it takes real understanding of how the technology works and its limitations. The Cambium cnWave team has been extremely helpful with design and topology issues as we’ve gone along the way. While there’s always room for improvements, we’ve seen improvements to software tools that can be used to understand what may be causing performance issues.
Our networks are in Phoenix area, so can’t speak to snow issues. Phoenix doesn’t get lots of rain, but when it does rain, it is very heavy for a relatively brief period of time. We’ve not seen outages due to heavy rain or even strong winds with DNs on 20’ poles.
I’m happy to answer any specific questions either here or offline.


Thought I’d add a little more detail. The Terragraph protocol is very complex, especially compared to the traditional multipoint equipment we’re all used to using for years. First, it is fundamentally a GPS sync’d TDD mesh protocol. This means the ‘AP’ is the backhaul and you need 50/50 down/up TDD timing. This creates more complexity and design constraints than one might initially think. Second, it includes an IGP, Open/R, which is similar to OSPF in concept. Third, the routing protocol is IPv6 only, so something is needed for IPv4. Cambium has a built-in L2 bridge tunnel that works well, but tunneling everything back to the core makes a large network much more sensitive to broadcast/multicast traffic. Finally, combining these 3 can make troubleshooting issues more difficult.

1 Like

@DigitalMan2020 our company, MCSnet, has roughly 5000 cnwave nodes deployed and operating in northeastern alberta, Canada. I will firstly say that whatever those operators are doing, they are likely doing something wrong. I have been happy about my decision to use the cnWave and have not looked back…

I don’t know why someone would suggest that rain makes dn’s experience issues @ 100m. That’s not something we have ever seen. I would suspect there are other underlying issues creating the problem that was described. (example: client setup for short range in the lab but didn’t change setting during deployment, interference between nearby nodes etc… etc…)… Or their distance estimates are off, and they are actually >200m apart… we have found that pushing distances on this front very greatly affects robustness. It may work on fair weather days… And, sometimes these poor shots can work as excellent backups… But I would never count on them as your only link

Now, we see occasional one off situations: wet sticky snow can cake a v3000 dish, the radomes help this situation but they don’t always work completely. The magic here is that it only rarely affects the network if you have multiple paths. If you design the network properly, the open/r will fail over and your clients won’t notice. Having multiple paths is a key element in our own network design.

In the case of our deployment, substantially all the radios are situated on subscriber premises. We are counting on them not to turn off the radios - and we have ways to encourage them to do so. But power outages or (very localized power outages) are a real occurrence. So we plan the network to ensure multiple paths and minimize any potential outage. It works surprisingly well and built out networks are fiercely reliable.

As we have grown, we have learned the power of this meshing feature. In one community, we had the privilege of migrating a cluster of subscribers from 900 mhz (nestled in trees) – they had a completely cluttered rf path to the tower, and they now get gigabit speeds thanks to multiple neighborhood paths.

cnWave is certainly a complicated system to setup, and I urge operators to learn to frequency plan and to understand how to use the e2e controllers etc. But once deployed, the system is capable of incredible performance and scaling.



I really appreciate everyone’s comments and the time taken to respond. My deployment possibility is a small village that we have a tower deployed already offering surrounding rural farms and acreages 25mbps to 75mbps and we have a 450m offering the town from 25 to 100mbps. We would really like to jump into the wireless fiber speeds game and offer the village this opportunity. The deployment does have some hurdles as there are a good amount of trees, not too dense, but some larger poplars/rowed trees. This and the already lack of interest from the village residents to switch to us from the local dsl provider who can only offer 50mbps, on a good day lol. Having invested in the 450 lineup and it not panning out has me hesitant to invest in the cnwave lineup. But maybe the residents are awaiting for someone to offer fiber like speeds.