Faster 802.11 WLANs In The Works: A Progress Report

WIRELESS LOCAL-AREA NETWORKS (WLANs) based on the IEEE 802.11 Ethernet standard continue to enjoy growth and success. Such WLANs are widely deployed in large and small organizations, homes, and a growing number of public access points called "hot spots." Data rates have increased from the 802.11b's standard 11 Mbits/s to 54 Mbits/s with the newer 802.11g and 802.11a standards.

These WLANs, known as Wi-Fi for wireless fidelity, are essentially standard on virtually all new laptops and many PDAs. Wi-Fi also is finding its way into some cell phones. Current standards are more than adequate for routine LAN e-mail connections or Internet access. Now, attention has turned to establishing a new version of the standard that can address the growing interest in, and need for, higher WLAN data rates.

IEEE TASK GROUP N Late last year, the IEEE established Task Group n (TGn) to create a standard that would provide a minimum data rate of over 100 Mbits/s. Development for this standard, designated 802.11n, is moving forward rapidly. Dozens of full and partial proposals from semiconductor and equipment companies as well as academia are being considered.

As usual, the standards process is messy and contentious as organizations fight to get their ideas and intellectual property included. Luckily, most of the visible proposals have common elements that should enable a new standard to emerge in a reasonable time.

Those who have worked on this problem agree that the best solution involves orthogonal frequency-division multiplexing (OFDM) and multiple-input multiple-output (MIMO) spatial diversity and spatial division multiplexing to achieve the 100-MHz+ data rate over a respectable distance.

OFDM divides a signal into many lower-speed bit streams and transmits them in parallel on multiple adjacent orthogonal channels. It's a DSP technique implemented with the fast Fourier transform (FFT) and the inverse FFT. OFDM also is already widely used in DSL, where it's known as discrete multitone (DMT), and in the 802.11a and 802.11g WLAN standards. The new fixed wireless broadband systems are turning to OFDM as well.

MIMO employs multiple transmitters and receivers to transmit parallel data channels on the same band. Using two separate transmitters and two separate receivers (2-by-2 MIMO) doubles the data rate. Also, the transceivers can take advantage of the diversity offered by separated antennas. This ultimately helps mitigate multipath interference, which is the bane of microwave transmission. Even higher data rates can be achieved by using three or four separate transmitters and receivers.

CHANNEL BANDWIDTH The new standard must be backward-compatible with current 802.11a/b/g equipment. This assumes the use of the current 20-MHz wide channels. Versions for the 2.4- and 5-GHz bands are being considered. If the bandwidth can be expanded to 40 MHz, even higher data rates are possible. Currently, 40-MHz channels aren't permitted, but they may emerge in the future. This becomes a particularly good option for the 5-GHz 802.11a band, where 440 MHz of spectrum space is available.

Already, the standards-setting process has developed two competing factions. Both employ OFDM and MIMO, but they differ in how these techniques are implemented.

One group, World Wide Spectrum Efficiency (WWiSE), comprises companies like Airgo Networks, Bermai, Broadcom, Conexant Systems, STMicroelectronics, and Texas Instruments. Its plan is to stay with the existing 20-MHz channels and use a 2-by-2 MIMO to achieve a 135-Mbit/s data rate. By going to a 4-by-4 MIMO scheme, up to 540 Mbits/s is possible in applications that can trade off higher speed for the greater expense of extra radios.

The other group, TGn Sync, consists of Agere, Atheros, Intel, and other companies. It proposes using 40-MHz wide bands that employ a 2-by-2 MIMO scheme to generate a data rate reaching 250 Mbits/s. With a 4-by-4 MIMO arrangement, the maximum rate could reach 500 Mbits/s. Speed and bandwidth go hand in hand, so this looks like the logical way to go, since higher speed is possible at lower cost.

The key is getting approval for the 40-MHz bands. This could happen in the U.S., but in other countries like Japan, it may be unlikely. Such a version would need a feature that lets the radios adapt to the 20-Hz channels and deliver a lower speed.

BETTER THAN UWB? What do you do with a 100-Mbit/s+ WLAN? The goal is to eventually sell the 802.11n radio modems to the consumer electronics market, where demand is growing for audio and video transmission between TV sets, cable boxes, DVD players, audio systems, camcorders, and other devices. Range is expected to be between 15 and 20 meters for a 100-Mbit/s rate. That's much farther than the 10 meters expected from the forthcoming ultra-wideband (UWB) products.

UWB is much further along in the development cycle, even though no one standard has been set. It appears that two versions of the IEEE 802.15a UWB standard will coexist for now. Both versions, one using OFDM and the other using traditional direct-sequence CDMA-type UWB, will be available early next year. Each will be able to achieve the 100-Mbit/s+ rate.

As the new 802.11n standard is finalized, it could overtake UWB due to its longer range and potentially higher rate. Look for the standards work to continue for now. Final ratification isn't expected until mid- to late 2006. The first products are potentially slated for early 2007. To track the progress on this work, go to grouper.ieee.org/groups/802/11.