Question : Problem: Does increasing the length of the passphrase of WPA - TKIP decrease throughput?

Hi, I am currently experimenting to determine and measure the overhead/throughput decrease that WPA Personal - TKIP encryption causes to wireless networks. the test bed I used is a Linksys AP along with 2 very similar laptops (1 wired-100mbps ethernet and 1 wireless - 802.11g). I use the software called LanTraffic on both laptops. On the wireless laptop, I generate packets sizes ranging from 1024KB-8192KB and measuring throughput with a 1ms inter packet delay. I then user encryption - tkip: a 10 letter passphrase and repeat the above tests and then keep increasing the passphrase length by 10 every cycle.
I tried it using 54Mbps at first and then decreased it to 6Mbps and get different numbers but again no signifcant reduction.

The weird part is that I do not get any significant reduction in the throughput. Is this normal? Does anybody know why this is going on? or what I can change to get more accurate results? Any help will be appreciated

Thank you in advance

Answer : Problem: Does increasing the length of the passphrase of WPA - TKIP decrease throughput?

The passphrase used in WPA is only used during the initial connection of the 2 wireless devices. It doesn't have to do with the encryption key itself. Once the 2 devices are communicating they use self generated encryption keys which they change constantly so that it makes them more difficult to collect enough data or crack. The key length is defined internally by WPA and I don't know what gets used.

The passphrase should just be long to make it more difficult for 3rd parties to guess. If they guess that passphrase, they could get access to the router, but they won't be able to crack the traffic between the router and other wireless connected users.