- 10 open source companies to watch
- Mythbuster busts his own tale
- $208 million petascale computer gets green light
- Sony recalls 73,000 Vaio laptops
- Chrome and Firefox and add-ons
Newsletters | Podcasts | Chats | Opinions | RSS Feeds | This Week In Print | IT Careers | Community | Reports | Downloads | Slideshows | New Data Center
Partner Sites:App Performance | On Demand Security | Networking Solution | SOA | Value of WDS
Linksys this week announced a line of 802.11g gear that the company says significantly boosts wireless LAN speeds without the use of proprietary technology. SpeedBooster, an extension of the 802.11g standard, increases performance by 35% on a SpeedBooster-only network, and up to 20% when used with standard 802.11g products, Linksys says.
The move is in response to the success small office/home office network rivals Netgear and D-Link Systems are having with Super G, a wireless chip technology developed by Atheros. Super G uses a proprietary technique called "channel bonding" to achieve 108M bit/sec-rated speeds when used with similar equipment.
Netgear and D-Link started selling Super G gear last year; Netgear says Super G makes up about 30% of its 802.11g equipment sales. Touting a standards-only approach, Linksys has resisted and seen a slight drop in WLAN market share - from 56.59% in December 2003 to 53.13% in January 2004 - to sales of competitor's faster equipment.
Chip vendors such as Atheros and Conexant have developed higher-speed wireless products powered by a mix of standards-based technology and proprietary techniques such as packet bursting, hardware encryption, and most notably, channel bonding. The development is because the IEEE's 802.11n standard, which is expected to get 100M bit/sec speeds, is at least a year away and to meet the performance demands of emerging entertainment networks.
However, such products only achieve higher speeds when they communicate with other similar products, which flies in the face of the Wi-Fi Alliance's push to ensure all Wi-Fi products interoperate. The group has certified products using Super G and others for interoperability, but only in standard 54M bit/sec mode. The Wi-Fi Alliance says it will not certify any vendor's 108 mode or any proprietary mode.
D-Link and Netgear use the same Atheros Super G chip, but whether their products interoperate isn't clear. D-Link says they do. Netgear says its customers say they do but Netgear won't support interoperability with D-Link Super G. Atheros says they interoperate; rival Broadcom, which developed the technology on which SpeedBooster is based, says they do not.
But the biggest problem with Super G isn't the lack of interoperability: It's the use of channel bonding. Network World columnist Kevin Tolly of The Tolly Group conducted tests (which Broadband commissioned) in December that showed standard 802.11g networks suffer severe performance degradation when near a Super G network. 802.11g has 11 channels, but only three are non-overlapping, which means you can operate three wireless networks in the same area without interference, so long as they are set to channels 1, 6 and 11, respectively. To achieve higher throughput, the technique "bonds" together Channel 6, half of Channel 1 and half of Channel 11, degrading nearby networks' performance.
Last week, Tolly's team completed a new round of tests, extending the range between Super G and standard 802.11g networks to 30, 50, 100 and 150 feet. Tolly also tested an Atheros Super G network against an Atheros standard 802.11g network. These tests address claims that the first tests placed the networks too close together and that testing Super G against Broadcom standard 802.11g equipment created a bias.
Although the results won't appear on Tolly's Web site until April, he gave Network World a sneak peek. In testing a Netgear Super G network against a Netgear standard 802.11g network positioned 30 feet apart, the mean was 48.5M bit/sec for the Super G network, but 1.4M bit/sec for the standard network.
"The new results don't show anything that contradicts what we found previously," Tolly says. "Even at 30 and 50 feet we found significant interference, which represents what users in condos and apartments could experience through walls, ceilings and floors. Many lots aren't 50 feet wide, and there are plenty of places where you don't have 50 feet between you and your neighbor's DSL connections."
Tolly's tests also found that placing two Super G networks near each other degrades performance because both must use Channel 6, of particular significance to users who buy a second to expand the network.

This comprehensive, 115 page guide provides frontline network troubleshooters with practical advice...
File Integrity Monitoring: Secure Your Virtual and Physical IT EnvironmentsDiscover the capabilities your file integrity monitoring solution should have to effectively secure...
Realizing the Potential of User-Generated and Social NetworkingCan communication service providers (CSPs) leverage Web 2.0 services and create new service...

The standard for Power over Ethernet (PoE), IEEE Std. 802.3af(tm)-2003, advanced networking,...
Intelligent Mobility: BlackBerry Technical Seminar 2008The virtual BlackBerry Technical Seminar keeps growing in popularity every year, and we want to...
Harnessing the power of communications to increase workplace performanceDue to the convergence of IT and telecommunications technologies, the business workplace has been...

WAN Ethernet services are reliable, cost-efficient offerings that are widely available and in a...
Comments (1)
How to boost a wireless connection?By Anonymous on May 28, 2007, 9:12 amHi I'm not getting a good signal from our WLAN. how can I strengthen it in my notebook?
Reply | Read entire comment
View all comments