I was reluctant to link to this article because the subject of the article violates my credulity: Now, I could be wrong. I am not an expert on either information theory or low-level physical layer limitations for radio. But the folks at Extricom are offering something that I believe is impossible while discussing it in the vaguest possible terms.

The company's Web site is incredibly thin on their actual technology. They have a good white paper explaining the technical limitations of current technologies, and the realities of re-using channels over short distances. They claim that you can only reuse a channel in the 2.4 GHz band three times over 1,200 feet without suffering some bandwidth degradation. They're probably right.

But their technology is undefined. They say that they have solved the problems that restrict frequency reuse and talk about using each channel at triple its capacity: that 33 Mbps for 802.11b for a total of 99 Mbps across three channels, or 162 Mbps for 802.11g or 486 Mbps for three channels. They would bond 802.11a channels for 1.296 Gbps capability.

They say that their TrueReuse patent-pending technology enables this miracle, but they don't say how in even the slightest way. They allude to the fact that there are a lot of wasted time slots on a normal Wi-Fi network in either band dealing with packet collision avoidance, and in the intervals when devices back off after a collision. Fair enough. Entirely true. But they don't really say how their technology could triple the throughput of each channel, nor how they can avoid harmonic interference on the side nodes of seemingly nonoverlapping channels in 2.4 GHz (1, 6, and 11 do overlap, just not very much).

I'm in a wait-and-see mode. The Web site describes compatibility with existing Wi-Fi technology, but the article about the technology at Wi-Fi Planet quotes the company saying that all Extricom technology is needed to achieve the highest bandwidth--I confirmed with the reporter that that means all cards, APs, and switches. It will work at slower rates with existing Wi-Fi devices.

Another strange item on their Web site: their switch is described as being the single product for an entire network. Can multiple switches interact to create a larger-scale network? The FAQ describes adding APs for scalability, which won't work when the switch is full. The switch has a single 100/1000 Mbps connection, which is inadequate for a modern network--especially when the company is talking about offering 1.296 Gbps from one switch. A device of this scale needs, at a minimum, two high-performance switched gigabit Ethernet interfaces--and possibly more than that.