airmax quality/capacity

[drwho17]

Is there any documentation on what these represent, or how they are calculated? Trying to get a handle on some throughput issues I'm seeing, where the air rates/ccq/amc/amq don't seem to correlate to the actualy throughput (same channel, same AP).

[avolve]

Same here, I just got my sector and NanoM5's and am starting to test out the Rocket5M with the Nano5M's.

[RebusCom]

I would also like to know what the AirMax Capacity spec. actually represents. We have a 19km PtP link with two 5M bullets running v5.1. Config. is AP-WDS to Sta-WDS, autonegotiate rate @ 65/65Mbps, 20MHz bandwidth, AirMax on, NoAck, CCQ is 100%, AirMax quality is 98-100%, but AirMax capacity is only 47%.

Measured throughput using Jperf is 25Mbps running three streams, which is about half of what I would expect. We have a regular 20MHz 802.11a PtP link hopping from the same location that negotiates at 48/48 that Jperf measures 20Mbps throughput across.

[cwnetwork]

I would also like to know what the AirMax Capacity spec. actually represents. We have a 19km PtP link with two 5M bullets running v5.1. Config. is AP-WDS to Sta-WDS, autonegotiate rate @ 65/65Mbps, 20MHz bandwidth, AirMax on, NoAck, CCQ is 100%, AirMax quality is 98-100%, but AirMax capacity is only 47%.

Measured throughput using Jperf is 25Mbps running three streams, which is about half of what I would expect. We have a regular 20MHz 802.11a PtP link hopping from the same location that negotiates at 48/48 that Jperf measures 20Mbps throughput across.


You must probably already know this I hope;
Airmax is polling , polling is not necessary for PtP, the extra overhead makes it slower in PtP situations ....WDS is also used in situations where more than one AP is present ... thus not PtP as well..... one drawback of WDS are that throughput is generally halved when enabled but in your case with only one ap and one station this presumably does not apply.
The advantage you can have with M units in PtP is that is 802.11n, which modulates to higher bandwidths.

[RebusCom]

You must probably already know this I hope;
Airmax is polling , polling is not necessary for PtP, the extra overhead makes it slower in PtP situations ....WDS is also used in situations where more than one AP is present ... thus not PtP as well..... one drawback of WDS are that throughput is generally halved when enabled but in your case with only one ap and one station this presumably does not apply.
The advantage you can have with M units in PtP is that is 802.11n, which modulates to higher bandwidths.
Airmax must be enabled for NoAck operation. The link was also tested with no AirMax and the throughput changed only about 1Mbps -- very little difference.

[avolve]

I have a rocket ap sector setup with two nanom5's and notice Air Capacity goes down when you change the Airmax setting from none to high. Setting it from high on down through the list does affect it's bandwidth. However both rocket and nanom5 tell me I am connected at 200megs and the rocket is connected directly to a switch that is connected directly to a server with a speedtest on it. It rarely goes over 21 megs, consistenly closer to 20 megs when AirMax is on and set to highest. Setting it to none tests out higher than low. This is with only two Nanom5's and really no traffic from one of the nanom5's.

[davey]

> WDS is also used in situations where more than one AP is present ... thus not PtP as well

A PTP link that is used as a transparent bridge MUST use WDS and there is no performance hit.

[drwho17]

You must probably already know this I hope;
Airmax is polling , polling is not necessary for PtP, the extra overhead makes it slower in PtP situations ....WDS is also used in situations where more than one AP is present ... thus not PtP as well..... one drawback of WDS are that throughput is generally halved when enabled but in your case with only one ap and one station this presumably does not apply.
The advantage you can have with M units in PtP is that is 802.11n, which modulates to higher bandwidths.
AP-WDS - Station/WDS does not halve bandwidth, and I'm not sure your understanding is very clear on what it does or how it is used in a p2p situation, WDS must be used for transparent layer2. Airmax is recommended for PtP it doesn't make a difference in throughput on UBNT gear. The advantage isn't necessarily 802.11n, the advantage is airmax uses TDMA, no CSMA backoff.

[gunther_01]

He is probably used to WDS repeater like modes. Which do cut speeds in half.

[drwho17]

He is probably used to WDS repeater like modes. Which do cut speeds in half.
Sure, but that's not what the poster said he was using.

[gunther_01]

True......

[RebusCom]

To return to the topic of the thread, does anyone know what AirMax Capacity represents? It seems to correspond to our lack of throughput, but that may be coincidental.

[UBNT-Mike.Ford]

To return to the topic of the thread, does anyone know what AirMax Capacity represents? It seems to correspond to our lack of throughput, but that may be coincidental.

Hello,

AirMax Capacity is a factor of data rates + errors. If your data rates are lower then they could be at a given signal then the Capacity will be low.

1x1 devices (AirGrid, bullet) can not have over 50% capacity.

Thanks,

Mike