On the surface, the wireless throughput numbers look fine, with about 4Mbps performance in all four test locations. I think part of this, however, is due to the excellent 802.11b section in the NETGEAR WAB501 Dual-Band card [reviewed here] that I used as a wireless test client.
When you look at the Ixia Chariot plots shown in Figure 2, however, a different picture emerges.
Figure 2: Wireless Performance
(click on the image for a full-sized view)
The plots confirm the good average throughput numbers, but you can see a consistent pattern of wide variation in throughput, under both strong and weak signal conditions. I also experienced long delays before many of my Response Time tests completed (note the Condition 2 results) that seemed like the AP went to sleep for awhile. This effect was pretty consistent in all test locations, and is probably linked to the same mechanism that is causing the throughput swings.
On a postitive note, WEP-enabled throughput was virtually indistinguishable from non-WEP operation, although it was difficult to tell for sure, given the wide throughput variation.
802.11b Wireless Performance Test Results
|Test Description||Signal Quality (%)||Transfer Rate (Mbps)||Response Time (msec)||UDP stream|
|Throughput (kbps)||Lost data (%)|
|Client to AP - Condition 1||96||4 [No WEP]
4.1 [w/ WEP]
|Client to AP - Condition 2||62||3.9||278 (avg)
|Client to AP - Condition 3||69||4||3 (avg)
|Client to AP - Condition 4||55||3.8||3 (avg)