Wireless Access

I just bought a AP-135 to learn and test with.  I have a Netgear WNDR4000/N750 (2  yr old) at the house also.  I needed to do testing for an outdoor deployment that was going to cover a 200 foot distance.   Everything was the same, APs within a foot of eachother, same switch and internet conection, same MacBook Pro client. 

The capabilities of the Aruba over the Netgear don't even NEED to be discussed.  It is obvious.

However, even when within 10 feet of the APs, the Netgear consistently had better throughput.  About 15-25%.  The Airport utility showed that when on the SSID of the Netgear the transmit was steady at 450mbps.  And my download throughput showed that stability.  Switch over to the Aruba and the transmit dances from 270 to 407.  I rarely saw it flick to 450. The throughput also oscillated from 60%-90% of the netgear's. 

I started with FairAccess, switched to DefaultAccess, and then PreferredAccess, thinking that might have been it.  Throughput did appear to improve under PreferredAccess, but it always lagged noticeably behind the Netgear.

Any settings or gotchas with 450mbps performance that I might look for?  I didn't expect to have lower throughptu being all alone on the AP.  I do realize the AP may, over time, make radio adjustments but I assumed this throughput would be a slam dunk in this clean and clear situation.

I just updated the AP to the available code this weekend.

The client is : 

en1:

  Card Type:AirPort Extreme  (0x14E4, 0xD6)

  Firmware Version:Broadcom BCM43xx 1.0 5.100.198.104.5)

  Supported PHY Modes:802.11 a/b/g/n

  Supported Channels:1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 36, 40, 44, 48, 52, 56, 60, 64, 100, 104, 108, 112, 116, 120, 124, 128, 132, 136, 140, 149, 153, 157, 161, 165

The 450 is just the data rate realistically speaking your wireless card will not be able to consistently sustain a data rate at 450  since it will depend on the health of your RF environment among other things :

- How many APs are using that same channel in the area

- Spatial streams supported on your wireless card

- The amount of clients connected to that AP

- How far the device is from the AP

- SNR

plus you have to take in consideration that the wireless network is shared medium 

If you need to obtain the actual throughput you use iPerf or  jPerf.

Yes, I get that.  But isn't 450 *ONLY* acheivable through 3 streams?  With my download host I saw about 15-16mbps throughput with a single download. I could have loaded up a series and maxxed my internet connection(50m), but the single one provided enough consistent result that I was satisfied. 

Secondly, this is an apples to apples test. The same client (using the airport -I cmd to get instantaneous info), as I said, WAS maintaining 450 as the last transmit rate with the Netgear, but did not with the Aruba.  

As for channels, that was the responsibility of the Aruba to select, measure, decide and adapt.  That's the advantage of the product.  I was always on an A channel--which is much clearer.  The Noise levels were -9# to -8# on A channels.  It was considerably worse and varied on B/G.  And as said in the post my distance in my testing was five feet to 200 feet, and the comparative results were always the same. 

I was the only wireless client on both, after all this was a test.  What would be the point of random load?

On your controller do a show ap arm rf-summary ap-name <apname> and show ap arm neighbors <apname> this will give you an idea if there more devices using those channels .

ARM will adjust but it also depends of how dense is your deployment or how close is that AP to your other Aruba APs and since you are using 40 MHz the amount of channels available are less too.

Is your Netgear 4x4:4 by any chance ?

Don't know what you mean about random load ?

I'l set the two APs up later today again and try those commands.  Have a client to visit. I do know the number of interfering APs went down as the aruba kept running, which I assume was it finding clearer channels. 

The netgear is two years old.   I'm fairly certain it is a 2x3.  I don't think Netgear has a 4x4 in their lineup even today. 

This was a test, in a residential (single dwellings) neighborhood with grassy islands and no trees (clear line of sight to the AP).  There are no random clients involved, either the AP had 1 client(me) or none.. This was a test.  One macbookpro client moving back and forth from the aruba to the netgear.  The APs were a foot away from eachother so the density of the environment is the same (apples to apples), but quite low on the A side of thing.  I only know of three A radio APs on my part of the street, and I own own two of them.