How We Tested
My wireless tests have always been done by the walk-it-around method, so are as flawed as anyone else's. So when Azimuth asked whether I would consider testing draft 11n products using their system, I thought it as good a time as any to try a different test methodology.
Azimuth's approach puts both the AP and STA in RF-tight (they call this RadioProofTM) boxes. The enclosures come in different sizes and port configurations and typically provide 90 dB of isolation over 1 to 6 GHz. The enclosures include filtered power and data connections to prevent RF leakage from those sources. Putting the products under test in RF shielded enclosures essentially removes the problem of WLAN and other interference sources.
Once in the enclosures, the products under test were connected through programmable RF attenuators, which for MIMO testing have a 60 dB range and are used to attenuate signals to test "range" effects. Direct, cabled connection was made to all three antenna connectors on the AP and STA radio cards as shown in Figures 10 and 11 below. Connection was made in each case between antennas in the same card physical position (left-to-left, center-to-center, right-to-right).
Figure 10 shows the Linksys WRT-300N connected up for test in one of the test enclosures. You can see red and green Ethernet cables connecting to two LAN-side ports since Broadcom-based products don't include gigabit Ethernet ports. So instead, for some tests I used the same two-Ethernet-client technique as in the RangeMax 240 review so that the 100Mbps port speed wouldn't limit throughput test results, which are supposed to exceed 100Mbps.
Figure 10: Linksys router in test enclosure (Click to Enlarge)
Figure 11 shows the Linksys WPC-300N Notebook card in an IBM / Lenovo T42 ThinkPad with a 1.7 GHz Pentium M and 512 MB of RAM. The Notebook was running WinXP Pro SP2 and Azimuth client software.
Figure 11: Linksys card in test enclosure (Click to Enlarge)
The table below summarizes the card driver and router firmware revisions tested.
|Product||Chipset||Driver or Firmware Revision|
|Linksys WRT-300N||Broadcom Intensi-fi||0.92.6|
|Linksys WPC-300N||Broadcom Intensi-fi||22.214.171.124|
|Buffalo WZR-G300N||Broadcom Intensi-fi||1.42 (1.0.37-1.07-1.03)|
|Buffalo WLI-CB-G300N||Broadcom Intensi-fi||126.96.36.199|
|Netgear WNR854T||Marvell TopDog||1.1.05NA|
|Netgear WN511T||Marvell TopDog||188.8.131.52|
|Netgear WPNT834||Airgo Gen 3||1.0.41|
|Netgear WPNT511||Airgo Gen 3||184.108.40.206|
In each case, I checked the manufacturer's website and downloaded and installed any newer firmware that I found. I should note that the only updates I found were for the Linksys WRT-300N and Netgear RangeMax 240 routers.
Also of note is the situation that I found with the Buffalo product. After I ran the tests on the Azimuth system, I received review samples directly from Buffalo. I found that the 220.127.116.11 driver supplied on the CD was newer than the 18.104.22.168 driver that came with the retail-purchased product. I then checked the Buffalo download site, but did not find the driver posted.
When I checked with Buffalo, I was told that the driver that I received was now included in currently shipping product and was available by calling Buffalo customer support. After further discussion, Buffalo ended up posting the new drivers for download the same day we spoke.