Bench testing Wireless Bridges

I was looking to create a bench testing procedure for wireless bridges so that I can compare bridges from different manufacturers in a controlled environment. I was wondering if you experts have any advice or best practices for doing this.

What minimum distance should I have between the antennas in order to make the test worth anything?

Are there any software programs written that will push traffic across the "link" and analyze the performance?

What criteria should I test for?

Your thoughts would be greatly appreciated. Thanks for looking!

I normally bench test bridges across the length of the workshop (about 4m), often pointing the antennas away from each other so the RSSI is around -50dBm as some radios get upset when picking up too strong a signal.

For thrashing the network for throughput testing I use a free bandwidth testing utility from Mikrotik, and rely on the reporting capability of the bridges to check for errors, dropped frames and other stats based on connection reliability.

Taking this a step further would be to insert attenuators in line with the antennas to simulate reduced RSSI and see how the bridge performs under such conditions.

There are factors that are difficult to test without a large distance between the antennas due to the ACK timeout, environmental issues, etc. so ultimately be aware you will find the link may behave slightly different once set up regardless of how it gets thrashed on the bench.

__________________
Jeremy

Marlborough WiFi Project

Hi there

I often test Wifi bridges for throughput and this is my method. I normally set the APs about 20 metres apart and set up the bridge. I run a program called NetPerSec which shows through put on the fly and keeps records of max and min levels. I used to do a FTP transfer from the desktop to laptop but found no difference from just doing a file copy from a networked drive. I will often run a continuous ping as well to check how well the network performs under pressure.
In the real world you will find the performance may vary greatly due to interference and also the 802.11b protection feature that most APs have that will reduce throughput greatly.
Throughput can also vary with encryption type as some require more data bits. The mode used will also effect throughput ie. bridge - bridge or AP - client mode or WDS - WDS.
Cheers
Dale