Wireless LANs move to 5 GHz

he wireless LAN market is coming on strong. Last year approximately 7 million units of WLANs were sold, adding up to an estimated $1 billion market. That’s impressive, but WLANs have yet to realize their full potential. Systems built to a new IEEE standard, 802.11a, will soon appear to take advantage of more frequency spectrum, a more favorably crafted set of FCC regulations and more advanced modulation techniques.

IEEE 802.11b allowed wireless networks to achieve Ethernet-equivalent speeds. This and the establishment of the Wireless Ethernet Compatibility Alliance (WECA) as the industry forum for ensuring Wi-Fi (Wireless Fidelity) interoperability and the decision by major notebook makers to integrate WLANs into mobile PCs have helped to ignite the WLAN market. Table 1 shows how new 5-GHz spectrum regulations and IEEE standards provide 802.11a systems with a solid foundation for growth. These include more frequency spectrum, higher data rates and more advanced modulation techniques. The resulting significant benefits to 802.11a users are likely to be very significant and should be carefully studied and understood.

The serious concern over interference associated with 2.4-GHz WLANs is warranted because potentially large amounts of consumer electronics (such as cordless phones) and Bluetooth devices are being released into the market to the angst of 2.4-GHz WLAN users. This is the result of very liberal guidelines for transmit power levels in the 2.4-GHz Industrial, Scientific and Medical band. Within this band, FCC 15.247 rules allow any frequency-hopping or direct-sequence spread spectrum device to have a maximum peak output power of 1W. There is no regard for separating low-powered devices (such as Bluetooth and cordless telephones) from high-powered ones (such as fixed broadband wireless-access systems) to different parts of the spectrum so that they do not interfere with each other.

For the 5-GHz Unlicensed National Information Infrastructure (UNII) band, FCC regulations 15.407 are more expertly crafted. Table 2 shows different output power requirements for different parts of the 5-GHz spectrum. (Maximum ratings refer to peak output power delivered into the antenna. Equivalent isotropic radiated power measures the power transmitted by a directional antenna in the strongest direction.) The intent of these rules was to separate and mitigate potential interference between different applications.

For example, the upper UNII band from 5.725 to 5.825 GHz would be best suited for outdoor Fixed Broadband Wireless Access devices that typically require much higher power to reach longer distances. Likewise, the lower 200-MHz band would better serve shorter indoor (as well as outdoor) WLANs in the 5.15-to-5.35-GHz band. Along with those total power limits, the FCC has also specified power spectral density limits, which force systems with bandwidths narrower than those covered by 802.11a to transmit with less power. Furthermore, all UNII devices must be high-data-rate communication devices. This implies that previously mentioned 2.4-GHz narrowband interferers (that is, cordless phones, low-rate Bluetooth devices) are not likely to find a home in the 5-GHz band.

The 802.11a systems use a modulation technique known as orthogonal frequency division multiplexing (OFDM) to mitigate multipath effects. In real environments, multipath exists between the receiver and transmitter and occurs when the transmitted radio signal is reflected from walls, furniture and other indoor objects. Under such circumstances, the transmitted signal may not have a single direct path to the receiver. That’s why models, which rely on radio wave propagation characteristics in free space, do not accurately form real 5-GHz performance.

Furthermore, real environments exhibit a number of different paths, or multipaths, each of which has a different distance to travel from the transmitter to the receiver, thus experiencing a different delay. As a result, a transmitted signal can have multiple copies of itself and arrive at the receiver at different moments in time. Thus, from the receiver’s point of view, it receives multiple copies of the same signal, each with a different signal strength or power.

This can be visualized with the help of Fig. 1. Here, T represents the time interval between successive multipath signals. The maximum amount of time between the first and last multipath signal at the receiver is known as the delay spread, or tmax. The ratio of tmax/T denotes the maximum number of copies of the same signal that can be received. The larger this ratio, the less multipath resistant a system is.

OFDM decreases a 802.11a system’s sensitivity to multipath by using a unique parallel transmission scheme. Instead of transmitting information using one frequency, or carrier, once every time interval T, OFDM divides the total transmission data among N different frequencies or subcarriers. With respect to each subcarrier, the amount of data transmitted has been decreased by 1/N. Or equivalently, on each subcarrier the transmission interval has been lengthened by N for the same amount of data.

However, despite the fact that the data rate for each individual subcarrier has been reduced by a factor of N, the paralleling of N total subcarrier transmissions means that the overall transmission rate of the system still remains the same. This means that the tmax/T ratio, with respect to each subcarrier, has been decreased to tmax/(T x N). Each subcarrier is now N times more multipath-tolerant. For IEEE 802.11a systems, the number of subcarriers is equal to 52. The technique is the basis for 802.11a’s superior multipath resistance.

CMOS has proved to be a remarkably resilient technology, one that has been improved at every technology generation. As a result of these efforts, CMOS devices are fast enough for use in building high-frequency integrated circuits. One way to quantify this increase in speed is to examine the ft, or current gain cutoff frequency, of mainstream CMOS process technologies. This cutoff frequency metric represents the highest clock rate that can be achieved with any given technology generation. In almost all cases, the continuous shrinkage of the minimum gate length of a CMOS device has led to ever increasing cutoff frequencies. For CMOS technologies with 0.25-micron gate lengths, the ft value is between 20 and 30 GHz–more than enough for 5-GHz applications.

Another cost saver is the ability of CMOS to allow integration of many functions on a single chip, which is not possible for SiGe and GaAs due to yield and manufacturability issues. Integration not only brings overall silicon costs down but provides savings in chip testing, sorting and packaging as well. Atheros has chosen to integrate all 802.11a radio, baseband and MAC functions into a two-chip solution using 0.25-micron digital CMOS technology. The design (Fig. 2) eliminates external components such as external VCOs, SAW filters, PA and LNA as well as external MACs and memory common in other WLAN solutions.

These are exciting times for the WLAN industry. The arrival of 802.11a brings to the market never before seen wireless LAN performance. The advantages of 802.11a over current WLAN solutions are significant. These are the result of, among others, a combination of increased spectrum, additional high data link rates and FCC transmit power rules. Additional performance gains can be realized with OFDM for more robust links in multipath prone environments. These advantages coupled with low-cost, high-volume CMOS implementations allow 802.11a WLANs to penetrate new markets and increase their penetration of the enterprise, home and public "hot spots." Get ready: WLANs are poised to soar to a whole new level with 802.11a.