06-25-2013 08:55 PM
I recently deployed a controller based LAN mesh for a client on a 3400 running 18.104.22.168 and using AP-134 with directional antennas.
It was a fairly solid link path over about 100m using 2 x AP-ANT-93.
When the link came up ,i noticed that when i ran a "show ap mesh topology" command, the link speed was maxing out at 300/300 exactly.
Now whilst i expect some loss and associated speed drop, i was at least expecting the link to fluctuate between 300-450 MHZ (HT mode was enabled and this link only used 5GHZ). No matter how we aligned the antennas or altered the gain settings, 300 was it.
this was a concern as my client had chosen the AP-134 over the 104 and similar for the 450Mbps speed.
I did some tinkering and found a setting in the radio profile for the number of spatial streams assigned for STBC on TX / RX. This was set to 1. I figured being a direct LOS and point to point link, i didn't really need any logical diversity so i turned this from 1 to 0 .
My thoughts were (correct or not i don't know) that this SS was used for diversity and hence thats why the link was topping out at 300Mbps.
Changing this made no difference, even after both ends of the mesh were rebooted.
So my question is, can an AP-134 sustain a 450Mbps link over Mesh or is there some other undelying issue at play here?
06-25-2013 09:34 PM
As you suspect, there is a deeper issue at work.
3SS indoor APs like AP13X and our new AP22X depend on multipath from the walls, ceilings, floor, objects and even people in order to recover 3 distinct spatial streams (and corresponding 3SS MCS rates).
Outdoors we do not have much if any multipath. So we create multipath using dual polarization antennas. The separate H and V polarities are "distinct" enough that it allows the radio to recover two different streams.
Unfortunately, there are only two such polarizations to work with. You didn't mention which antenna you have, but if it is one of our 3 element panels you really only have 2 polarizations. If you look carefully at the arrows on the back, two of the arrows are at right angles to one another. The third arrow is at 45 degrees to the other two. But if you do the vector math, it turns out that the 45 degree one is just partially H and partially V. There is no "third" polarization.
There are such things as circular polarization, however these don't help us outdoors because the E and M fields created by the circular element just ping pong between H and V.
Bottom line, it's really not possible to achieve 3SS rates outdoors without significant local structures that would create enough indirect multipath for the radio to recover the 3rd stream.
In other words, this is a basic property of radio.
To achieve higher speeds outdoors, you need to increase your channel width to HT40, though this exposes you to more interference potentially.
06-26-2013 09:38 PM
Thanks for the explanation, i was suspecting as much however i think there might be more to it yet again.
My client upgraded thier controller to 22.214.171.124 and now the link is showing up at 450Mbps which is quite strange. might be a query for the TAC..