Over the next few years wireless local area networks (WLAN) will migrate to the 5 GHz band, which offers better prospects for higher data rates along with potentially fewer interfering communications. One would hope that after all the fragmentation in the 2.4 GHz band, particularly 802.11b vs. HomeRF, the industry could unite behind a single global 5 GHz WLAN standard. Think again.

We appear poised for a yet another round of industry fragmentation, only this time, thankfully, the differences between the two technologies are not as severe as between the direct sequence 802.11b and the frequency hopping HomeRF.

This time the competing technologies are IEEE 802.11a and HiperLAN2.

Technology Overview

The IEEE 802.11 committee specified 802.11a for WLAN systems operating in the 5 GHz band. This specification is based on orthogonal frequency-division multiplexing (OFDM) and will allow data rates of 6 to 54 Mbps.

The European Telecommunications Standards Institute (ETSI) has developed the High Performance European Radio Local Area Network 2 (HiperLAN2) standard. It too features OFDM for the physical layer (PHY) and specifies data rates up to 54 Mbps.

However, this is where the current similarities end. While 802.11a is essentially implementing an Ethernet-like scheme, HiperLAN2 defines an Asynchronous Transfer Mode (ATM)-like architecture that is particularly suited to voice and multimedia applications. HiperLAN2 is arguably the more technologically complex and superior of the two, implementing a more sophisticated media access control (MAC) layer.

In addition to the provisions for isochronous communications, which provides an inherent quality of service (QoS), HiperLAN2 implements Dynamic Frequency Selection (DFS) and Transmit Power Control (TPC), as required by European rules. DFS helps reduce interference and allows for better utilization of the spectrum, while TPC helps adjust the power output level.

Global 5 GHz Frequency Bands

Note that unlike the 2.4 GHz band, the 5 GHz band is far from uniform around the world. Different regions and countries have allocated different blocks of spectrum for next-generation WLAN services. Moreover, there are different rules governing different bands, even within the same country. For example, in the US while the 5.15-5.25 GHz and 5.25-5.35 GHz bands are adjacent they are subject to different rules on power output.

The lack of a uniform global 5 GHz frequency band poses a challenge to both silicon vendors and OEMs. It is far more difficult to sell a global market solution.

Who's Got "The Big Mo"?

HiperLAN2 could be described as a primarily European effort being pushed by telecommunication companies such as Ericsson, Nokia and Siemens, designed to provide a seamless communications network between 3G cellular networks and private networks.

In contrast, 802.11a has the support of the much of the WLAN community, particularly those vendors currently selling 802.11b solutions. Of course, there are a number of players with one foot in both camps prepared to support both technologies if necessary.

Most 5 GHz WLAN silicon vendors are focused either only on 802.11a or on delivering 802.11a chips before HiperLAN2 solutions. Hence 802.11a has greater momentum, and will reach the market before HiperLAN2. 802.11a products will be available in late 2001, while HiperLAN2 devices are expected sometime in 2002.

It would appear at this early stage that though HiperLAN2 is arguably the superior technology, 802.11a is poised to dominate. 802.11a has time-to-market and momentum on its side, which are critical success factors in the WLAN market.

However, this does not mean that HiperLAN2 is doomed to failure and will disappear. On the contrary, HiperLAN2 will have a strong presence in Europe and a number of major consumer electronics companies, such as Sony, are closely evaluating HiperLAN2 because it is optimized for multimedia networks.

802.11a is Evolving

However, 802.11a is not standing still, and the IEEE is addressing its limitations. The 802.11h working group is developing DFS and TPC enhancements so that 802.11a solutions can be sold in Europe. Also the 802.11e working group is developing QoS enhancements to the 802.11 MAC.

Thus 802.11a will become more full-featured over time, and though its technological underpinnings may differ somewhat from HiperLAN2 it will incorporate many of the same features.

The promise of these enhancements is a critical element in ensuring that the bulk of the WLAN community is behind 802.11a rather than HiperLAN2.

Convergence

There have been multiple efforts to converge key aspects of 802.11a and HiperLAN2 into a single global 5 GHz WLAN standard. However, these have largely been unsuccessful. And with 802.11a products expected to start shipping in late 2001, such efforts face a tremendous uphill battle.

However, some silicon vendors have begun suggesting that with advancements in flexible digital signal processing (DSP)-based architectures, dual-mode 802.11a/HiperLAN2 solutions are not only technically possible, but also economically feasible. These dual-mode chipsets would essentially implement a very flexible MAC solution and leverage the commonality of the OFDM PHY.

This approach would be a boon to global OEMs that don't want to have to customize their WLAN-enabled products for different markets.

A Last Couple of Points

In the final analysis though there will be a repeat of WLAN fragmentation in the 5 GHz band it will be mitigated by the fact that HiperLAN2 will predominantly be adopted in Europe and that silicon vendors will drive toward dual-mode solutions in the longer run.

As an additional point it would be prudent to note here that the migration to 5 GHz will take time because backwards compatibility with 2.4 GHz technologies remains an issue, as do questions about the cost and range of 5 GHz solutions.

Allied Business Intelligence Inc is an Oyster Bay, NY-based technology research think tank publishing research on communications and emerging markets. Details of these studies can be found at www.alliedworld.com  or by calling 516-624-3113 for more information.