Analysis A year ago, many were writing off 802.11a in the home market, claiming that 802.11g had won the battle and that the 5GHz variant would only carve out a market in the enterprise, if at all. However, the ‘a’ strand of Wi-Fi is biting back in the digital home as frustration with interference problems in the overcrowded 2.4GHz band mount, and in the past week, both Linksys and Netgear have announced dual-band products that work in both frequencies simultaneously.

This signifies not just that 802.11a will carve out a niche after all, but that it stands a strong chance of dominating the areas of Wi-Fi that carry the highest margins and are therefore most attractive to the innovative vendors. As such, we could see a swing against 802.11g, especially as chipmakers like Atheros start to work round the limitations of 802.11a in terms of range, and the problems of overcrowding in the 2.4GHz band become ever more threatening.

Markets for 802.11a

There will be three key sectors where 5GHz Wi-Fi will flourish – outdoors, the enterprise and the multimedia home. In the first of these, its most natural home, it will ironically face the greatest obstacles – from WiMAX, 802.11b mesh and regulatory complexities around outdoor 5.8GHz. In the enterprise, we believe 802.11a has a fairly clear run, given the low penetration of other WLANs and the cautiousness of corporate buyers in moving towards newer developments such as WiMAX or pre-standard 802.11n. (The most talked about market for Wi-Fi, indoor internet access, is the least likely to boost upgrade moves from 802.11b because of the limitations on typical backhaul in home and hotspot sectors.)

In the home, 802.11a is up against the large installed base of 802.11g, and is facing the emergence of products running at over 100Mbps, usually operating in 2.4GHz (and labelled as either ‘standards-plus’ 802.11g or ‘pre-standard’ 802.11n). There is also the nascent challenge from technologies that are optimized for fast multimedia, notably 802.15.3 or WiMedia.

The belated success of 802.11a in these three sectors will depend on three factors – the level of real user frustration with 2.4GHz interference, the pricing of ‘a’ equipment, and the performance comparison with alternatives. The attitude of large influencers like Microsoft and Intel will also play a part. If users start to see real benefits to investing in ‘a’ in terms of cleaner spectrum and specific applications like video streaming, there will be a genuine market demand for the first time and an increasing awareness of the different advantages of the two 54Mbps Wi-Fi standards. It will then be up to the chipmakers to ensure that they achieve the pricing and performance to meet this demand in an attractive way.

Will that user demand materialize? In the enterprise, the answer is almost certainly yes, now that dual-band products are available that can protect any existing investment in 802.11b. Enterprises full of Bluetooth gadgets and Wi-Fi laptops will be keen to avoid interference problems and to take advantage of other advantages of 802.11a in terms of real world performance. Plus, its shorter range, and the resulting denser pattern of APs required, is less problematic in an enterprise setting.

Outdoor market

Outdoors, ironically, given that this was seen only recently as 802.11a’s main haven, is where there will be the biggest question marks. This is partly because of the cheaper alternative of mesh for unwiring large districts, and the promise of pre- WiMAX gear for WISPs or enterprises seeking to cover a metro or campus aream with high bandwidth services.

There is confusion in the Upper Band of 5GHz (5.725-5.825GHz) - in some countries the only piece of spectrum allocated for outdoor Wi-Fi - because WLAN allocations overlap with the ISM frequencies (5.725-5.850GHz), which allow a far wider range of devices, raising the threat of interference from ISM devices to 802.11a signals. For instance, some DECT cordless phones are being developed for the ISM range.

The Upper Band overlaps with 802.16 profiles and other broadband wireless services as well and has a higher 4W EIRP, which has not been verified in all countries (for instance, India, parts of Europe, Korea). This band is allocated to fixed wireless as the primary service in India, Korea, Australia, Sri Lanka, Malaysia, Japan, Thailand and Vietnam and so is likely to be dominated by WiMAX rather than Wi-Fi.

This is not to say that there will be no place for 802.11a in the outdoor market – companies like Vivato have made a strong contribution to getting around its limitations in range using smart antennas and other techniques.

The digital home sector

But the real prize will be a mass home market. According to recent research from Parks Associates, home networking will be a $12bn business in the US alone by 2008, while In-Stat/MDR is more cautious, predicting $17.1bn globally by the same date, split about equally between the US, Asia and EMEA. Both companies believe the Microsoft XP Media Center and the PC will be the key drivers and that Wi-Fi will be the primary network.

The defensive driver behind 802.11a, as we have seen, will be dissatisfaction with performance of 2.4GHz technologies. With a huge number of home devices, from microwaves to baby monitors to Bluetooth handsets, operating in this spectrum, the potential for interference degrading performance to unacceptable levels is high, especially when users are looking for quality-sensitive applications such as VoIP and video streaming.

There is far less risk of such problems in 5GHz, partly because there is far more spectrum available (only 83.5MHz of 2.4GHz spectrum for WLANs, compared to 455MHz and, in some countries, more, for 5GHz WLANs). And, outside the ISM bands, apart from incumbent radar devices, it is reserved just for WLANs. Also, the technology of 802.11a is more spectrally efficient because its is OFDM-based – a more efficient approach than

802.11gt’s direct spread spectrum, and incorporates support for interference mitigation techniques (see box). Because 802.11a WLANs support as many as 24 non-overlapping channels, they are less susceptible to interference than 802.11b or 802.11g.

As well as interference issues, 802.11a is better able to cope with other difficulties in the home environment, such as severe multipath channel impairment, with subsequent negative impact on performance, from the large number of in-home reflections and obstacles. OFDM avoids many of the channel impairment issues, though the higher frequency requires stronger antennas to penetrate walls and to achieve long ranges.

Such devices are starting to be developed and to enter the mainstream, and the development of CMOS radios, and more compact 802.11a chipsets and antennas, will narrow the cost difference with the 2.4GHz Wi-Fi variants, and make the technology more appealing to consumers. Advances connected to chips and to OFDM will also help to mitigate 802.11a’s native disadvantages, compared to 802.11g, in terms of range and antenna size.

New applications

A positive driver will be needed, however, to make consumers decide en masse to stop putting up with the quirks of 2.4GHz and turn to a new generation of equipment – or for consumer electronics equipment vendors to make the investment of incorporating 802.11a, or dual-band, as standard. This driver will be superiority of 802.11a for multimedia, high bandwidth applications, particularly video.

Though in its early stages, the demand for in-home wireless video and audio networks is expected to mushroom from 2005 onwards and 802.11a is increasingly being seen as the best basis for such networks, at least before 802.11n and UltraWideBand-based 802.15.3 are widely available – a window of around two years, possibly longer if the standards committees working on both these specifications fail to achieve a united proposal.

Many experts feel that Wi-Fi is being stretched to its limits by digital home applications, being forced to support speed, quality of service and other requirements that were never envisaged for a wireless extension of the common Lan.

This is why, we believe, the winner in the second generation digital home market will be UltraWideBand (whether based on IEEE or de facto standards) with fast DSL or WiMAX backhaul. But for the early adopters, 802.11a is the best currently available solution, offering higher real world speeds than 802.11g, better quality of service, better indoor coverage and greater spectral efficiency.

Chipmaker moves

Home equipment suppliers are waking up to this fact, and they are being encouraged by the chipmakers’ moves to make 802.11a more optimized as well as cheaper. Intel’s eventual announcement of its tri-mode Centrino, which will ship in PCs and consumer electronics devices early next year, was a major boost to the market and Atheros, in particular, has been at the cutting edge of extending 802.11a’s range, data rate and price/performance.

Even more significant in the digital home market is Toshiba’s work on a chip based on a variant of 802.11a technology, optimized for home entertainment equipment.

The Japanese giant is working on a version that would work with devices based on the vanilla standard but would make it easier for televisions and other non-PC equipment to receive video and other multimedia files.

Toshiba has created chips using the new specification and will incorporate them in some of its own consumer equipment, though it would not give details on availability to other vendors beyond saying it would submit its developments to the IEEE.

A success for 802.11a would, in the short term, benefit one company more than any others – Atheros, which provided the silicon for the new Netgear dual-band range. Atheros bet its marketing strategy and many R&D dollars on the standard in 2002, only to see Broadcom crashing into 802.11g before it was even standardized and achieving a Wi- Fi market lead almost overnight.

Since then, Atheros has focused on superfast Wi-Fi silicon for all three variants, and on ever more compact chipsets, targeting the digital home and consumer electronics markets, but it is still looking for a significant payback on its extensive 802.11a R&D, and a chance to leapfrog Broadcom again.

Linksys and Netgear

The moves by Linksys and Netgear show that the connoisseurs of the consumer WLAN market are starting to take 802.11a seriously. The lack of 802.11a client products has held back dualband products – those supporting 5GHz for 802.11a and 2.4GHz for 802.11b/g in one device – but this situation will change as the interest in 802.11a’s strengths for multimedia applications intensifies. With dual-band, users can connect simultaneously to both networks, relying on ‘a’ for video and ‘g’ for email and surfing.

Netgear’s new range is labelled ‘Double’ to reflect the dual bands and the boost in speed and range from the Atheros chipsets – up to 108Mbps data rate from its SuperA/G technology and range of 400 feet through walls using its XR enhancements. The Double is the first consumer range to incorporate XR, which was announced earlier this year. Its first two products are the Double-108Mbps Wireless Router, priced at $129 with WPA security and firewall, and a $79 PC Card. A USB adapter is due later this year.

Meanwhile, Linksys’ dual-band offering is Wireless A+G, which includes the WRT55AG router Linksys has had since 2003, though with a significantly lower price, down from $299 to $109. It has also launched a new notebook PC Card and desktop PCI adapter for $89 each and an external USB adapter for $99. All the products are targeted at users of multiplayer games and video streaming, the strongest applications for 802.11a.

The real breakthroughs will come when Linksys and Netgear start to incorporate such devices into their home multimedia gateway products, and still more, when the likes of Sony and Toshiba are embedding 802.11a in their products. 802.11a will be dependent on the expected boom in home multimedia networks materializing in the 2005-6 timeframe, if it is not to be a footnote in Wi-Fi history. But so far, its chances are looking strong.

Copyright © 2004, Wireless Watch

Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.

Further background

802.11h

The ITU has laid down common conditions for sharing spectrum with incumbents such as military radar in 5GHz bands. These rely on two agile radio techniques for reducing interference, Dynamic Frequency Selection (DFS) and Transmit Power Control (TPC). These are enshrined in the 802.11h standard extension, which also helps mitigate interference between different WLAN devices and which defines mechanisms that allow 802.11a devices to comply with the ITU’s recommendations.

The main significance of 802.11h is that it enables vendors to create a ‘world mode’ 802.11a product that addresses interference rules in all major markets. Similar mechanisms will be devised for 802.16. While the main objective is to avoid conflict with incumbent users such as the military, the 802.11h standard and its 802.16 equivalent will also help reduce interference between all unlicensed and/or secondary use equipment.

The 5GHz bands

In 1996 the FCC opened up new unlicensed spectrum - 350MHz in the 5GHz bands at 5.15-5.35GHz and 5.725-5.875GHz, for use by a new category of unlicensed equipment, called NII/SuperNet devices and mainly concerned with fast wireless Lans. This band became known as the UNII band. Canada followed with virtually identical rulings.

The new band would avoid the free-for-all of the 2.4GHz spectrum by restricting usage to wideband communications radios (though in 1996 there was no conception of how widely such devices would be used by a decade later). Other countries began to make similar rulings and at the ITU World Radio Conference of 2003 (WRC-03) there was a significant move to global harmonization of 5GHz open bands, their usage and anti-interference mechanisms. At this conference, the ITU laid down global recommendations that allocated the 5.47-5.725GHz band for WLANs, following European precedents for HiperLan, in addition to the 5.15-5.35GHz frequencies, a total of 455MHz of new global spectrum for WLANs.

Unlike in 2.4GHz, the ITU effectively locks out competing devices from the 5GHz WLAN bands, although it requires that WLANs do not interfere with primary users such as radar and satellite. But WLANs also get primary status, which comes with rights of protection from interference from any future users of these frequencies that the ITU may consider.

Some countries are considering additional spectrum in 4.9-5.0GHz, notably Japan, or above 5.825GHz, notably the US, which has reserved 75MHz of spectrum in 5.85-5.925GHz for its Dedicated Short Range Communications Service, aimed at roadside and vehicular communications and to be based on 802.11a.

However, the Wi-Fi Alliance may not certify 5GHz devices below 5.15GHz or above 5.85GHz, at least until after 2005, and some countries like Japan and India have not opened up all the available bands recommended by the ITU. Also, countries vary in which sections of the band they reserve for indoor or outdoor usage – Europe, for instance, has yet to harmonize policy on outdoor usage in the Upper Band.

The ITU had also, in 1997, declared some new bands for international unlicensed usage, which overlapped with UNII, but were for different purposes. These are for unlicensed ISM usage (industrial, scientific and medical) and, unlike UNII, are not restricted to a particular type of radio. The most relevant ISM bands to WLANs are 2.4-2.5GHz and 5.725-5.875GHz. The latter is open in most countries for commercial fixed access wireless communications and overlaps with the Upper Band of UNII. This overlap is the most serious source of confusion and potential interference for 5GHz WLANs going forward, especially in the outdoor market.

It is worth noting that the ITU’s ISM band (5.725-5.85GHz) is open in many countries for WLANs, but is not officially part of the ITU allocation for WLAN. This Upper Band is not uniformly available across Europe or in Japan and European nations will roll it out at very different and often unspecified paces, while Japan has not made a final decision on opening it up at all. This upper band is mainly relevant to outdoor 802.11a systems delivering hotzone or broadband wireless access services, and to 802.16.

Related stories

Linksys sends corporate WLAN standard home
Intel unveils tri-mode Wi-Fi for Centrino
Agere refocuses on Wi-Fi