ARTICLE JULY / AUGUST 2002

Strong market demand for wireless LANs

Wireless LANs (WLANs) are in their infancy. Current networks are relatively small and inflexible and cannot replace wired networks. The first interoperable WLAN technology - IEEE 802.11b - showed a strong market demand for applications in companies and home offices. However, it also identified security, scalability and ease-of-use problems.

With HiperLAN2, ETSI BRAN therefore set out to provide:

• excellent security;
• high/scalable capacity;
• managed bandwidth with predictable performance for each user;
• quality of service levels to meet the needs of any type of service; and
• auto-configuration tools to simplify use.

And this is supported by robust protocols to optimise throughput to make the most of radio bandwidth at 5 GHz - the 5 GHz spectrum offers well-defined channellisation, power control and dynamic frequency selection, and is relatively insensitive to interference.

The approach taken by BRAN is based on cellular or ad hoc networking. There are two basic modes of operation:

1. Centralised - this uses cellular topology with access points controlling radio cells each covering a certain geographic area. This approach allows mobile-to-mobile and mobile-to-base system communications and is aimed principally at business applications, where the area to be covered is larger than a single radio cell.
2. Direct - this uses ad hoc networking in a single radio cell, typical of a private home. Here mobile terminals within the cell can exchange data directly.

BRAN's HiperLAN2 standard is highly flexible to enable it to be used with a variety of core networks. The architectural platform defines a core-network-independent physical layer and data link control layers with a set of core-network-specific convergence layers on top of the data link control layer.

The physical layer provides protocol mapping and uses orthogonal frequency division multiplexing (OFDM) signal modulation. OFDM was selected as it supports high data rates with very low packet and delay losses over distributed wireless networks. Low latency and great flexibility are ideal for real-time mobile interactive and multimedia applications. The result is higher quality service and greater cost effectiveness than with other current wireless technologies.

The data link control ensures basic data transport functions and radio link control for exchanging data between the access point and the mobile terminal. Its interface uses time division duplex (TDD) and dynamic time division multiple access (TDMA). And, finally, the convergence layers adapt service requests from higher layers to the services offered by the data link control, and converts higher level data packets into the fixed-size data units used within the data link control.

The MEDEA+ A105 Universal wireless local area network (UniLAN) project is now developing and demonstrating the building blocks necessary for a common wireless terminal compatible with HiperLAN2 and the Bluetooth wireless standard. The consortium includes chipmakers, research institutes, equipment suppliers and service companies. They are looking initially at system architectures to achieve high data rates with a good quality of service and high security. They will then concentrate on developing the building blocks concerned - including RF modules as well as baseband digital signal processing and media access controller chips able to function with both Bluetooth and HiperLAN2.

UniLAN sees a major market developing for wireless office LANs in major enterprises - this sector is expected to represent more than 50% of the market by 2005. The consortium also expects wireless access points to be installed at airports, conference centres, hotels and caf?s. These would allow users to read and send emails, surf the Internet and carry out a wide range of LAN office applications from laptop PCs and other mobile terminals. The resulting broadband WLAN technology could also have applications in the commercial aeronautics sector to allow aircraft to exchange data with air terminal computing systems while taxiing on airport runways.

At the chip level, the MEDEA+ T204 ASGBT project is continuing development of advanced SiGe bipolar and BiCMOS technologies for mixed signal on chip wireless applications. BiCMOS is fast becoming the preferred technology for mixed RF-analogue chips while SiGe offers important cost and power savings in high-speed devices. The resulting chip technology should be equally applicable to transmissions rates of 2.2 GHz for UMTS mobile phones, 2.4 GHz for Bluetooth and 5 GHz for HiperLAN2 applications, as well as being able to work at up to 20 GHz or more for future systems.

The ASGBT project is setting out to obtain high bandwidth performance in active and passive devices with very low noise levels and very low power consumption. After optimising processes used to develop 0.35/0.25 ?m SiGe bipolar and BiCMOS technologies from the earlier MEDEA T555 project, it continues with the characterisation of 0.25 ?m heterojunction bipolar transistor (HBT) modules based on SiGe carbon substrates able to operate at the 80 to 100 GHz required for future wireless and wired optical fibre transmission applications.