7. Wireless Performance

NOTES:
• No Link Quality or Strength readings were available
• Testing was done with a Linksys WPC54G card in AdHoc mode in a Dell Inspiron 4100 laptop running WinXP Home

Rather than lug around the WET connected to my laptop, I plugged one set to AdHoc mode into my LAN's switch and used a Linksys WPC54G card also set to AdHoc mode to do the testing. As both the table above and Chariot plot below show, performance was surprisingly poor.

Figure 5: Four-location throughput test results
(click on the image for a full-sized view)

Note that his pattern is similar, but worse than, the results I obtained when I tested the Linksys WRT54G and WPC54G (Figure 6).

Figure 6: Four Condition Throughput test - WRT54G and WPC54G
(click on the image for a full-sized view)

To test whether the WPC54G was the problem, I tried a few tests with a Belkin F5D7010 54g CardBus card and also tried forcing the Linksys card to 24, 11, and 5.5Mbps transmit rates while testing at my Condition 3 location. In each case, however, I obtained similar results, so I have to conclude that the WET's radio is the cause of the poor performance.

I also tried a few close-range (my test Condition 1) tests with 802.11b clients.

Figure 7: Condition 1 throughput - NETGEAR WAB501 and Linksys WPC11v3
(click on the image for a full-sized view)

Figure 7 shows that the Atheros-based NETGEAR WAB501 performed nicely, but that the Intersil-based Linksys WPC11v3 had some problems with maintaining a connection. I retried the Linksys card about three times, and each time experienced connection drops before my one-minute run had finished. The NETGEAR card, however, was solid as a rock, even during a 5 minute extended test.

I also did quick tests at my Condition 3 location for both cards and found they maintained the throughput you see above. I guess the WET54G's Intersil-based radio is a better performer for 11b CCK-based operation vs. 11g's OFDM.

802.11g Wireless Performance Test Results

See details of how we test.