Forum

Hi folks,
I'm wondering if there's any way to test actual throughput on a WLAN? I'm guessing that maybe there's a software program out there that might do this?

I'd like to test the actual throughput between, for example, my workstation and the access point. I would be curious to see the difference(s) between the various 802.11b, g, and "Pre-N" equipment and the "theoretical" rates -vs- "actual" rates. :)

Thank you,
Craig

Hi Craig,

There are dozens of throughput measuring applications. I use netpersec, but there are many others.

The easiest way to do this is to use the WLAN client as an FTP client and put an FTP server behind the AP on the Ethernet network. Do a file transfer (using one big file). While this isn't necessarily representative of a real-world, multi-client BSS, it can show you the difference between use of various PHYs.

Devinator

Excellent! Thank you for your suggestions.

:)

Craig

I use software named InFielder by Motorola (formerly Wireless Valley).

I can hold a tablet computer at different heights, in corners, different orientations, each time clicking my exact location on a floorplan. I get a choice of udp or TCP throughput under realistic conditions of accidental shielding, and also get a better chance to identify areas of interference or especially bad multipath, or other factors that degrade throughput and would not be seen by just recording RSSI at the same locations.

Throughput will also show when the mobile device power or antenna isn't adequate, despite good signal levels from APs, or if there are client driver problems.

I used this method to compare standard 802.11g to MIMO performance, and found a definite performance difference with one version of MIMO.

I also use ftp and other software, but it takes a lot longer to collect location-based data points, since you have to make notes manually of the throughput at each location.

Although InFielder is the best tool I've found for this so far, it is too expensive for this type of experiment unless you already have a copy.

Charles Preston

Another tool you can use is IPERF.EXE , available on the internet.

I also use iperf. I find it to be simpler than the FTP example, as it doesn't require the installation of FTP client and server software (in fact, it doesn't require any installation at all, since it's command line) and it doesn't require procuring a big file to transfer. Finally, since it generates random traffic, it cuts the disk subsystem out of the performance analysis--not that the hard drive is going to be the bottleneck for an 802.11a/b/g wireless network, but maybe 802.11n?

Anyone that wants to do some SERIOUS stress testing on a WLAN environment might want to check out IxWLAN. This solution tests 802.11 Wi-Fi access points and the entire wireless LAN network by simulating multiple, concurrent virtual stations. It integrates with Ixia's performance testing applications - including IxChariot.

http://www.ixiacom.com/products/in_ixwlan

chariot is nice. plus, it'll spit out a detailed report to hand the customer.