Network Computing is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Analysis: Alternative WLAN Technologies: Page 8 of 16

Both Meru and Xirrus are sensitive to the power-draw issue. With so many radios, the maximum 15.4 watts of service that the IEEE 802.3af standard defines is insufficient. Fortunately, PowerDsine and other vendors are moving toward the next generation of PoE, currently nicknamed PoE Plus. The IEEE standard, 80.3at, will double--if not quadruple--power output and is due in mid-2008.

The last jab issued by critics of the wireless switch is real-world effectiveness and practicality. Are these AP designs too niche to enjoy mass-market adoption? Meru introduced the Radio Switch more than 18 months ago but only started shipping in mid-2006; according to the company, the school district of Philadelphia has purchased thousands of Radio Switches. And Xirrus announced a strategic alliance with ADC a few months ago and closed deals for 36 arrays, equivalent to about 276 APs. So at least some of your peers think not.

Conversely, the use of microcells by vendors such as Cisco and Aruba exacerbates co-channel interference, which drives up collision rates and degrades performance (see "Make Room for Microcells" page 70). Because DCF (distributed coordination function) rather than PCF (point coordination function) is the dominating medium-access method, clients can transmit whenever they sense the carrier is free. This lack of coordination works fine in low-density environments, but performance starts to stumble when utilization climbs. Also, in spite of microcells, clients choose where to associate, which means they can have less-than-optimal link rates.

Won't 802.11n solve these problems and marginalize vendors like Extricom, Meru and Xirrus? Unlikely. We're willing to bet that an 802.11a Meru WLAN with 100 clients will perform better than the first crop of enterprise 802.11n products with the same number of users. And by performance, we don't just mean aggregate throughput, but also jitter and evenness of throughput--the traditional 802.11n WLAN will have a broad bell curve of throughput rates, while the Meru system's curve will be relatively tight. By the time the market leaders have ironed this out, challengers will have 802.11n systems ready to go.

When In Roam