Hi all.
I have a work project which I want to research before throwing time and effort in to it. I’m hoping someone here can share their knowledge and experience.
I have a 4x GigE camera system. Currently, we connect to it over Gigabit Ethernet with MTU set to 9000.
We want to make it wireless and I came across the Mikrotik hAP ax2 devices. Our data rates hit around 90% of the GbE capacity (900Mbs). I am considering buying two hAP ax2 units; one connected to our GbE switch (Planet Networking), the other connected to our laptops’ Ethernet port. I plan to set the two routers up in bridge mode. They need to be 100% stable and operate over approx. 10m distance. We cannot afford any packet drops.
Can anyone here confirm if these devices are capable? Do they transfer GigE data without any filtering?
I’d appreciate any opinions before I purchase the hardware.
I do have a pair of those wireless wire devices. They work absolutely reliable for me since approx. 5 years. On 10m distance without significant obstacles they archive wire speed (in my case through 2 windows). Great devices.
In a nutshell, the Ax2's have omnidirectional antenna while in a PtP link the narrower the antenna, the better.
The (wireless wire) wAPs have a (still wide for a PtP) beam width of 60 degrees.
For higher distances much narrower beam width antenna are used and accurately pointing devices Is important.
But over 10 m distance anything will do.
The big advantage of 60 GHz Is that It gets much less interferences and "noise" than 5 GHz (let alone 2.4 GHz).
Over higher distances It may suffer from bad weather, but over 10 meters that shouldn't be and issue.
If you want a 5 GHz link, you can use two Sxtsq5 Ax, saving some money.
Thank you both for your replies.
The wireless-wire product does indeed look like the superior solution. I think we could manage to keep the receiver within the 60° beam angle. FYI. The sender device will be mounted on ah off-shore crane, while the receiver will be connected to a laptop. With some training, our technicians will be able to set up the equipment in the necessary orientations. The weather can get pretty gnarly, with high humidity and sometimes sea water spray. We can improve the IP rating. We will have to test for the effects on the 60MHz communication. But as you say, 10m should be okay. This is our (likely) maximum operating distance.
My only concern is the line-of-sight requirement and what would happen if someone were to walk between the two transceivers. I suspect the data will drop out instantly? I don’t know if our system will be able to recover quickly enough.
I will speak to our technicians and ask if this is likely to happen or can be managed.
Thank you again for your kind replies. I shall feedback on my decision and subsequent results.
Dr McNetface.
You must have a very short mast on that crane if someone can walk in front of it ....
I think we won't know the effects of that without some testing on the field, 60 GHz signals by their nature can be attenuated or blocked by very thin obstacles (very heavy rain or tree leaves) but since the distance between the devices is so small and the beam is so wide it may well be that the connection stays up during this transitions of people passing by because of reflections or whatever.
There is also the "big brother" of "wireless wire" that use two Cube Pro's that have two radios (a 60 GHz and 5 GHz) in a bond, so that if the 60 GHz fail automatically the (slower) 5 GHz takes over, but - loosely - the wAP are suggested for distances to up to 200 m or so while the Cubes for distances of up to 1000 m or so (and have a much narrower beam, 11 degrees, if I recall correctly), at 10 meter they might shout at each other so loud that I believe you would need to reduce their power.
In the context of @Netty_McNetface , the problem with most MT's WiFi6 devices (hAP ax2 included) is that they go up to theoretical 1200Mbps, while in practice they won't go beyond 800Mbps or so reliably and for actual payload ... because that 1200 number is about gross radio signaling speed and it's half-duplex while in reality there's significant loss in air-time for direction-switch, some 802.11 framing overhead including inter-frame gapping and due to some frame losses (due to interference/attenuation/noise/whatever).
So when thinking about a point-to-point link with certain (guaranteed) real-life throughput, one has to go for at least double theoretical throughput while considering actual causes for performance degradation. And considering limits of air-frame sizes (IIRC it's around 2340 bytes so clearly less than MTU sizes used in case described in opening post).
