802.11n: Already Too Slow?
June 07, 2005
Researchers around the globe are already looking at ways to get WLANs up to gigabit speeds.
Is 802.11n, the Wi-Fi technology promising to push wireless traffic to 100Mbps, already outdated?
"There is a common opinion throughout academia, industry and business that the current wireless technology fulfills neither current nor future demands," according to those behind the Wireless Gigabit with Advanced Multimedia (WIGWAM) project. This German-led consortium of corporate and university researchers says 100Mbps is the bare minimum needed for the future of wireless. It is using the 108Mbps 802.11n and MIMO technology as the starting point for bringing 1Gbps wireless into offices and homes.
Funded by the German government and supported by the likes of Siemens, Philips and Nokia, the project's goals are twofold: bringing wireless networks on par with wired nets, and meeting the increasing demands of a wireless world.
WLAN speeds will "increase to 100 megabits per second, at least," says Charlie Giancarlo, Cisco's Chief Technology Officer. The need for more range, more capacity and better quality of service are all pushing the need for more speed, according to Giancarlo.
WIGWAM "can be seen as 'beyond 802.11n,'" says project member Ralf Irmer. Irmer is the Vodafone Chair Research Associate at the Dresden Technische Universitat, the lead sponsor of the effort.
802.11n is viewed as "a good reference" by the WIGWAM project. The multiple in, multiple out (MIMO) technology it will use is a prerequisite to obtaining the 100MHz bandwidth using the 5, 17, 24, 38 and 60 GHz bands. While the project will end in 2007, it expects to see a 1Gbps WLAN by 2008.
Late last year, Siemens, a partner in the WIGWAM project, announced the first 1Gbps wireless connection. While slim on specifics, Siemens projected a tenfold increase in data and multimedia transmissions by the time the next generation of wireless is expected in 2015.
Although Irmer says the WIGWAM project has no intention of marketing any product, it has set out some basic hardware requirements. Materials for a base station unit should not exceed $100, and chipset, antenna and peripherals should fit into a PCMCIA, MiniPCI or Compact Flash II card.
"Individual partners of the project may market devices," says Irmer.
One WIGWAM partner, DaimlerChrysler, is working on ways to include 1Gbps wireless features in traffic and automotive applications.
WIGWAM's wireless 1Gbps development centers on three common scenarios: home and office, public access (hotspots) and high velocity mobile applications. Current WLANs do not meet the service quality and bandwidth requirements to replace enterprise wired networks. For the home, WIGWAM says, self-configuration, zero maintenance and low RF radiation are key.
"The high velocity scenario aims at railway and freeway information access," says Irmer.
The question of whether WIGWAM wants a new standard dealing with 1Gbps wireless is unclear. Although project members say WIGWAM doesn't create a full standard, the program's use of the 60GHz band falls within the 802.15.3.c millimeter-wave group formed in March, an offshoot of ultrawideband (UWB). WIGWAM's Irmer says influencing the standards process is important to marketing any 1Gbps devices offered by group participants.
The WIGWAM concept "will be brought into relevant standardization, like 802.11 or others," says Irmer. The WIGWAM project will be the topic of discussion at the July Wireless World Research Council meeting in San Diego.
Power consumption, energy efficiency and a WLAN MAC able to cope with the high data rates associated with 1Gbps wireless systems remain the group's key concerns.
Current WLANs "utilize only a fraction of the available bandwidth," according to the WIGWAM researchers.
The project is targeting the development of key components needed for Gigabit wireless to succeed, says Irmer. "For wireless 1Gbps, parallel vector DSPs with low power consumption are needed," he says. Effects from so-called "dirty RF," such as phase noise and jitter, need resolving.
WiGLAN, OGRE, Others
While WIGWAM is Europe's effort to prod progession of gigabit wireless, the subject is gaining interest around the globe. MIT's WiGLAN project is researching circuit designs allowing a 1Gbps wireless network using the 5.8GHz band. The MIT effort would create a small WiGLAN adapter connecting video streams, digital cameras, printers and mobile devices to the Internet. The adapter provides a 150MHz bandwidth using the 5.8GHz band.
In Australia, the government research arm CSIRO is working on using the 55GHz millimeter-wave spectrum to develop a 10Gbps wireless technology.
At the University of California at Berkeley, a group of researchers have teamed up to form OGRE: On-silicon Gigahertz Radio Exploration. The OGRE project focuses on the 7GHz of unlicensed spectrum available around the 60GHz band, an area key to WIGWAM and other wireless gigabit research.
Along with research, several startup ventures hope to cash in on the gigabit wireless concept. Isreal-based Extricom touts a combined Wi-Fi switch and AP tripling the bandwidth of 802.11a/g to more than 1Gbps. NewLANs is still promising 2Gbps to the desktop. Although NewLANs is active in the UWB community, the company's Web site remains blank.
While MIMO-based 802.11n may soon become a universal wireless standard, gigabit wireless is still some time off. Although the WIGWAM group predicts 1Gbps wireless systems by 2008, and Freescale Semiconductor expects to release its 1Gbps UWB chipset next year, a workable 1Gbps WLAN system is at best several years away, believes In-Stat analyst Sam Lucero.