Wireless Tests, Continued
But when I shared my results with both 3Com and NETGEAR, both said their product's intended use was to cover a room-sized area, such as a hotel or conference room. While neither company rep. said right out that the range was intentionally limited, neither seemed surprised by the question either.
A check of the maximum transmit power level test results in each product's FCC report confirmed identical lower-than-usual values of 17mW in 11b and 11mW in 11g modes - much lower than your average AP or wireless router's transmit power. I have to conclude that the transmit power was intentionally limited since the ASUS WL-330G mini-AP that uses the same Marvell chipset had FCC test results of 87mW in 11b and 191.9 mW in 11g.
The bottom line of all this is that both products should provide wireless coverage in their intended single-room environments. Sharing your connection with a buddy down the hall, however, will prove to be a frustrating exercise.
Figures 14 and 15 show how throughput is effected by engaging either 128 bit WEP or WPA-PSK / TKIP security. Although the NETGEAR fares better than the 3Com with WEP128 engaged (negligible loss vs. about 10% respectively), the NETGEAR loses about 35% and 3Com about 40% in throughput when WPA-PSK / TKIP is selected.
Figure 14: NETGEAR security mode throughput comparison
(click on the image for a larger view)
Figure 15: 3Com security mode throughput comparison
(click on the image for a larger view)
Although using WPA-PSK extracts a significant throughput penalty, once again, you need to consider the intended use of the products. The limiting factor in most cases will be a megabit-or-so Ethernet connection, so most users will probably never see a difference when using the recommended, stronger WPA-PSK. But since it appears that TKIP is straining the Marvell chipset, I doubt there will be a WPA2 upgrade, since that requires supporting more compute-intensive AES encryption.