One of the benefits of 802.11n is that it supports both the 2.4GHz and 5GHz radio bands. However, up until now, routers based on the still-in-draft 802.11n spec have not leveraged this capability. Apple is leading the charge to change all that with its Airport Extreme, the first draft 802.11n-compliant wireless router eWEEK Labs has tested that supports both wireless radio bands.
Apples AirPort Extreme exemplifies some of the performance benefits of 802.11ns dual-band capability, but it also highlights many of the new questions that wireless administrators must learn to ask as they evaluate the technology. The Airport Extreme is among the first draft-802.11n routers to utilize both the 2.4GHz and 5GHz bands, maintaining backward compatibility with both 802.11b/g and 802.11a.
In the last few months, weve seen numerous laptops with integrated draft-802.11n that support both bands (including the Lenovo T60p and Apple MacBook with which we conducted the tests for this report). To keep prices down, however, most wireless router manufacturers up until now have forsaken the 5GHz band, even though the wireless chip sets on which their products are based are capable of the feature.
In fact, along with the AirPort Extreme, the only other dual-band router weve heard of is the Buffalo Wireless-N Nfiniti Dual Band, which is in a different price league than the AirPort Extreme. At $179, the AirPort Extreme seems pricey for a wireless router, but not when you see the $299 price tag of Buffalos router.
That said, the Nfiniti N Dual Band has one big advantage over the AirPort Extreme: The Buffalo router supports both bands simultaneously. The AirPort Extreme, in contrast, has only a single radio, so we had to choose which band we wanted to support. An AirPort administrator configures the chosen band from the Airport console, and this choice is programmed in software on the radio. We have not tested the Nfiniti Dual Band, but Buffalo marketing materials claim it advertises both bands simultaneously to support different kinds of traffic concurrently.
Despite the AirPort Extremes relative shortcoming, we applaud Apples decision to support the 5GHz band. During tests of the AirPort Extreme, we found throughput performance to be significantly more stable in the higher band, due in no small part to the lower level of interference in the 5GHz band when compared with the vastly more crowded 2.4GHz band.
We also realized greater bi-directional throughout performance in the 5GHz band. In that band, our test MacBook topped out at 107M bps at close distances, while only achieving 90.5M bps in the same test run in the 2.4GHz band. Similarly, our Lenovo client performed about 22 percent worse at close distances in the 2.4GHz band than in the 5GHz band.
Interpreting the guidelines set forth by the 802.11n Task Group in its draft revision, Apple has not implemented channel bonding in the 2.4GHz band to help avoid interference with legacy wireless networks. This means that 2.4GHz-band 802.11n clients can use only 20MHz channels, while 5GHz clients can use a full 40GHz client. We could see this in each clients supplicant application: In 5GHz mode, each client would receive and transmit at the 300M bps link rate, while 2.4GHz clients could attach at only a maximum rate of 144M bps.
We also noted variable results in performance at longer distances in our tests. At 85 feet, the Lenovo clients performance dropped off 32 percent in the 5GHz band, compared to a 19 percent drop in the 2.4GHz band, while the MacBook dropped 7 percent in the 5GHz band, compared to a 28 percent fall in the 2.4 GHz band.
Although both of our client machines leveraged Atheros XSPAN draft-802.11n wireless chip sets (the same technology found in the AirPort Extreme router itself), we still noted a substantial performance discrepancy between clients. This indicates that, with the standards still in flux, consistent interoperability still remains a moving target—even within the same chip-set family.