Throughput vs. Path Loss
It's my practice to test wireless routers with their recommended client cards. Since Apple doesn't produce a companion Airport Extreme "Notebook" card for Windows machines, they sent along a top-of-the-line MacBook to fill the need.
But the Azimuth ACE 400NB Channel Emulator that is used for throughput vs. path loss testing requires the use of a Windows application on the wireless client and direct access to both router and client antenna terminals. I was able to get the Extreme opened up fine. But Apple frowned on my idea of opening up the MacBook, and after having checked out ifixit's excellent disassembly instructions, I have to admit I wasn't so hot on the idea, either.
I first tried to find an Atheros-based dual-band Cardbus card to match the Extreme's wireless chipset. But no vendor has yet shipped (or even announced) such a beastie. So after my direct appeal to Atheros for an engineering sample was politely refused, I had to switch to Plan C.
For 2.4 GHz testing, I used an Atheros XSPAN based D-Link DWA-652 Notebook card—the same used for the D-Link DIR-655 testing. The card was inserted into a Fujitsu P7120 Lifebook (1.2 GHz Intel Pentium M, 504 MB) notebook running WinXP Pro SP2 with all the latest updates. I used the 6.03.85 "Draft 2.0" driver and Windows Zero Config.
For 5 GHz testing, I used the only dual-band draft 11n CardBus card available—the Buffalo WLI-CB-AG300N used to test the Buffalo Nfiniti Dual-Band Router. The card was used in the same Fujitsu notebook with its 18.104.22.168 driver and Windows Zero Config. Note that since this card uses the Marvell TopDog wireless chipset, it doesn't necessarily represent best-case performance.
NOTE: We no longer refer to "range" in these plots, but instead use the more accurate "Path Loss". For an explanation, see the How we Test Wireless article.
Figure 26 was generated using the Compare Benchmarks tool of our Wireless Charts and shows 2.4 GHz up and downlink performance. Performance is surprisingly similar in both directions, which has tended to not be the case with most draft 11n products.
Figure 26: Throughput vs. Path Loss - 2.4GHz
Figure 27 shows performance in the 5 GHz band, which includes both 20 and 40 MHz bandwidth modes. Here, up and downlink performance is more like what I've seen with other products—noticeably different. As is typical with other draft 11n products, throughput falls off more sharply when running in 40 MHz bandwidth mode.
NOTE: 2.4 GHz and 5 GHz test results can only be compared within each frequency band. The Azimuth system does not reflect the difference in signal attenuation between the bands in the mathematical models it uses.