The 802.11n Standard: Grown Up at Last

High-throughput wi-fi standard promises to shape communications, especially the connected home

The 802.11n WLAN (wireless-local-area-network) technology is the only Wi-Fi technology today with the bandwidth to support multiple HDTV (high-definition-TV) streams at 20 Mbps each. This performance is sufficient for implementing some long-standing goals of Wi-Fi networks. One of these goals, wireless multimedia, comprises voice/VOIP (voice over Internet Protocol), data, video, and gaming in residential applications. Another goal is achieving throughput, QOS (quality-of-service), and security levels that compare favorably with those of Ethernet which are necessary for enterprise-grade, campus, and municipal networks. But the methods of achieving this performance are complex, leading to many of the options and variants in the standard. Another main reason for all the options is the large number of device types that users want to connect to Wi-Fi networks, each with its own distinct set of requirements. Because the market for Wi-Fi has become much more heterogeneous than it was in the early days of 802.11x, 802.11n networks must accommodate a much wider range of device types; many of the standard?s optional-requirement portions reflect that range. New requirements from consumer-electronics companies, such as video applications, or from the handset market, in which manufacturers are interested in power savings and better coverage, have contributed to the long, drawn-out IEEE process, as well as to the standard?s complications, says Frank Hanzlik, managing director of the WFA. "There are a lot more people to please, so the compromise process has been more complex," he says.

But now that WLAN capability has penetrated the consumer and communications-device markets and is increasingly embedded into DSL (digital-subscriber-line) and cable modems, as well as into Apple TV, the markets are bigger, and the stakes are much higher, increasing motivation and helping to drive resolution of those conflicts. As a result, 802.11n Draft 2.0 has turned into more of a framework than a standard, says Craig Mathias, principal of the Farpoint Group. How difficult it will be for engineers to navigate that framework remains to be seen.

What's in the standard?

The fact that products based on 802.11n can operate in either the 2.4- or the 5-GHz bands or both makes them potentially backward-compatible with legacy products. When you use 802.11n in the 2.4-GHz band and in a 20-MHz channel, it is backward-compatible either with 802.11b, using CCK (complementary-code-keying) modulation or with 802.11g, using OFDM (orthogonal-frequency-division-muliplexing) modulation. When you use 802.11n at 20- or 40-MHz channel widths in the 5-GHz band with OFDM modulation, it is backward-compatible with 802.11a. Devices built to conform to more recent 802.11x standards tend to employ OFDM because it is more efficient than CCK, which older Wi-Fi networks use. With 802.11g, OFDM has become the Wi-Fi-modulation scheme of choice. The 802.11n standard introduces more efficient OFDM modulation to increase data rate. It uses 52 data subcarriers instead of the 48 in legacy networks, producing 65 Mbps per spatial stream instead of the 54 Mbps of legacy 802.11a or g. Another option in the standard shortens the guard interval from 800 to 400 nsec to increase the OFDM symbol rate, further boosting the data rate.

Previous 802.11x standards specified only one frequency band, one channel width, one spatial stream -- transmitting or receiving -- per direction, and one maximum data rate. Aside from OFDM improvements to enhance throughput, the 802.11n spec also doubles the channel width and introduces frame aggregation, block acknowledgment, and spatial multiplexing; spatial multiplexing is one of several possible MIMO (multiple-input/multiple-output) configurations. It allows one or two channel widths, 20, 40, or both 20 and 40 MHz; one to four spatial streams in either direction; and at least two other MIMO options. Transmitting-data rate for 802.11n networks is therefore highly variable and is based primarily on modulation scheme, channel width, and the number of spatial streams.

Depending on the design, an 802.11n-compliant product can reach a typical throughput of 144 Mbps, assuming OFDM modulation, two transmitting and two receiving streams -- known as a 2Ă—2 configuration -- a 20-MHz channel width, or a (currently theoretical)maximum throughput of 600 Mbps, assuming OFDM modulation, a 4Ă—4 configuration, and a 40-MHz channel width. Most 802.11n products operating today achieve transmitting speeds between these two extremes: 300 Mbps with OFDM, a 2Ă—2 configuration, and a 40-MHz channel width, or 450 Mbps by simply changing the MIMO configuration to 3Ă—3.

As always, range also depends on several variables. But, with 802.11n, it's also more complex because of all the possibilities, including the transmitting power, number of receiving antennas, modulation scheme, and error-correction scheme. The spec also allows operation in 2.4- or 5-GHz frequency bands or a third dual-band 2.4/5-GHz option.

The differences between Draft 1.0 and Draft 2.0 are fairly minor, according to Bill McFarland, Atheros' chief technology officer and a member of the TGn. The biggest areas of change are associated with the coexistence mechanisms that govern how 802.11n devices behave with 802.11g devices, especially in the 2.4-GHz band. This issue is less important for 802.11a devices in the 5-GHz band: There, coexistence is simpler because the band is less crowded, and most devices perform OFDM-style transmission and exceptions. Also, the layout and organization of the frequency channels work well with 802.11n. "In contrast, the 2.4-GHz band includes 802.11b devices using CCK modulation, and the frequency channels are laid out in a way that doesn't match up as well with how 802.11n likes to use the frequency," he says.