Charlottesville (VA) - In some of the first comprehensive tests ever conducted on the third generation of wireless networking equipment, Tom’s Networking Managing Editor Tim Higgins revealed today, the performance of three so-called "draft 11n" WiFi routers exhibited surprisingly poor results. Using a sophisticated Azimuth Systems measuring device, Higgins tested new draft 11n routers from Buffalo, Linksys, and Netgear. Test results indicate that throughput rates for these draft 11n components dropped below that of a reference 802.11g WLAN - which presumably should have the lower speed - when the router (access point) is stationed no further away from the client than the distance of two typical household rooms.
Draft 11n equipment generally follows the standard set by the draft 1.0 document for the proposed 802.11n wireless networking standard, which was rejected early last month by the IEEE. WiFi equipment manufacturers have opted, for the meantime, to utilize chipsets that follow the rejected draft, rather than wait for the IEEE to ratify a draft 2.0 standard. Two of the routers in TNG’s tests - specifically, Linksys’ Wireless-N and Buffalo NFiniti - use Broadcom’s Intensi-Fi draft 11n chipset, while the third router, Netgear’s RangeMax Next Gigabit, uses Marvell’s TopDog draft 11n chipset.
Vendors of draft 11n equipment have touted throughput rates approaching as high as 240 Mbps, though engineers have privately warned that observed rates would likely approach something closer to 100 Mbps. In TNG’s tests, only the Linksys Wireless-N router was capable of sustaining an average throughput rate approaching 90 Mbps, with a signal loss within the two-room separation range. All three draft 11n routers in the tests managed to sustain average throughput ranges of about 85 Mbps within their own comfort zones, with the Linksys model managing to just eclipse 100 Mbps in the best-case-scenario test.
The TNG tests made use of an Azimuth W-Series WLAN Test system, in which the wireless routers and clients were placed in RF-shielded enclosures. Programmable RF attenuators were used to reduce the signal level between router and client, simulating the effect of increasing distance between the two.
The Linksys Wireless-N router began to show steep declines in average sustained throughput rate at a total path loss of about 74 dB, which Tim Higgins describes as proportional to two average sized household rooms or offices of separation. You don’t need to go out all the way to the back yard to see signals drop off, Higgins told us. Meanwhile, the Netgear RangeMax draft 11n model starts to exhibit dropoffs at about 72 dB, while the Buffalo NFiniti displayed a disturbingly finite dropoff point at about 60 dB.
Perhaps most surprisingly, the sustained average throughput levels of draft 11n equipment dropped below that of a recently tested Cisco 802.11g router (that’s a "g," not a typo) at a total path loss of about 80 dB. In other words, new draft 11n equipment may fail to keep up with existing 11g equipment when the transmitter is indoors and the receiver is outdoors.
For comparison, Tom’s Networking also tested Netgear’s RangeMax 240 router, which uses Airgo Networks’ True MIMO Gen3 chipset. As TG Daily reported on Tuesday, Airgo is going its own direction with regard to third-generation wireless equipment, opting to try to make its True MIMO architecture a de facto standard by means of widespread customer adoption. The RangeMax 240 did perform better than its three competitors in the draft 11n space, with performance dropoff starting as high as 81 dB total path loss, though its average sustained throughput rate failed to top 86 Mbps within its comfort zone.
These astonishing findings compelled TNG’s Tim Higgins to advise against draft 11n equipment purchases for the time being. "Draft 11n has a long way to go," he writes, "until it gets to a maturity similar to what 802.11g had when draft 11g products started to ship." Read the complete story, with detailed charts and test results for all four third-generation products tested, now on Tom’s Networking.