For some time, proponents of alternative wireless broadband technologies have been arguing that true high-speed mobile multimedia cannot be realised using wideband code division multiple access/frequency division duplex (WCDMA-FDD) technology. Companies such as Flarion, which is supplying the Flash OFDM (orthogonal frequency division multiplex) technology for the Vodafone trial, are keen to point out that the deficiencies of WCDMA-FDD in terms of latency and realizable data speeds basically make it an unsuitable delivery technology for wireless broadband.

This is hardly hot news. The GSM industry has been aware for some time that the FDD flavour of WCDMA which is being deployed to deliver 3G services does not have the horsepower to deliver on the data speed promises the industry has made.

'No budge' leads to fudge
The problem is historical. When the mobile industry foregathered in Paris in 1998 to debate which technology would be selected to provide 3G, there were a number of opposing camps supported by various groupings of operators and vendors. After much debate in the traditional smoke-filled rooms, the choice came down to an FDD or a time division duplex (TDD) solution. As no-one would budge, a compromise was agreed. The technology of choice for 3G would be a combination of FDD and TDD. The bottom line was that the new 3G technology deployed had to be backward-compatible with GSM, which remains the dominant second generation digital mobile technology.

WCDMA-FDD operates in paired spectrum in the 2GHz frequency bands allocated for 3G by the World Radiocommunications Conference. TDD operates in the unpaired spectrum. When the 3G licences were awarded the majority of operators acquired both FDD and TDD spectrum allocations as part of their licences. The strategy of operators was to push ahead with deploying WCDM-FDD networks, with the intention of deploying WCDMA-TDD overlays further down the pike as a solution for traffic hotspots. In addition, the development of WCDMA-FDD was more advanced - so it was logical to begin with FDD and move onto TDD.

However, this elegant plan was upset by the realization that, although WCDMA-FDD operating in paired spectrum is very good at supporting voice traffic, it is not so effective at delivering the high data speeds promised on the 3G 'tin'. Data traffic is inherently asymmetric; the uplink traffic tends to be much less than the traffic on the downlink, as any regular user of the Internet will confirm. WCDMA-FDD is a symmetric technology, which limits its capabilities as a delivery mechanism for the type of high-speed data services which operators were hoping would provide the new revenue streams they need.

GSM operators around the world spent more than US$100mn on acquiring 3G licences and are committed, at least in principle, to spending about as much again deploying their 3G networks. Every operator's business plan called for the deployment, in the first instance, of WCDMA-FDD technology. Changing technological horses in mid-stream is not a realistic option and in any event the majority of operators are mandated to deploy WCDMA by their licences. However, there is growing acceptance that operators must evolve their networks in some way if they are to meet the expectations of their subscribers in terms of data speeds. The question is: in which direction and with which technology?

Decisions, decisions
Unfortunately, from the operator's perspective, there are rather a lot of options. What about integrating WLAN into their networks to provide high speed data access in areas of high demand? Or wait for the arrival of HSDPA which promises higher data speeds? Could WiMAX be the answer? What about complementary solutions such as WCDMA-TDD? Or alternatives such as OFDM?

Some solutions such as WLAN integration can be dismissed. Although WLAN can deliver the data speeds there are issues over security and above all of mobility which is a key requirement of any planned enhancement to WCDMA-FDD. HSDPA could be the answer but it is at least two years away from deployment even without counting the almost inevitable teething problems encountered by all new technologies. This leaves the alternative broadband wireless access technologies, which can be broken down into a number of distinct areas.

The first grouping are technologies which are either already standardised or are close to standardisation within 3GPP and 3GPP2. These include TDD and time division/synchronous code division multiple access (TD-SCDMA)) and the main players in this arena are companies such as IPWireless and Interdigital Communications. The second grouping is focused on the IEEE 802.16 committee which deals with what are known as 'nomadic local area wireless solutions'. Specifically this is the committee which has developed WiMAX, whose major proponents are Intel and Nokia. Then there is the IEEE 802.20 grouping which is looking for a new air interface for IP-based mobile networks. The main technology under consideration is OFDM, which is being pushed by Flarion and Arraycomm.

Going down the TDD route is highly tempting for mobile operators. Firstly they already have the spectrum allocation for TDD operating. Secondly, the technology has already been standardised. Thirdly, it was always intended that FDD and TDD networks would work side by side so there are no issues of mutual interference as has already been proven in trials undertaken by IPWireless and Nortel. TDD is certainly the most cost-effective option as the same cellsites and antennas can be used.

Heavyweight backing
A number of players have thrown their weight behind TDD as is evidenced by the growing membership of the Global UMTS TDD Alliance. Launched at the 3GSM World Congress in Cannes in February 2004, the Alliance has grown to include 25 operators and 15 equipment manufacturers and vendors. New operators Caribbean Hub (Grenada), Cats-net (Tanzania), Cedarcom (Lebanon), Douala1 (Cameroon), EmilNet (Mozambique), Excelcomm (Indonesia), Showers (Nigeria), NGTel (Nigeria), Telecentra (Latvia) and Vimplecom (Russia) join founding members including Airdata (Germany), AtlasOne (Malaysia), IPMobile (Japan), IQ Networks (Australia), Kite Networks (US), Mobicom (Mongolia), Net2Cell (Ireland), NextWave Telecom (US), Sentech (South Africa), Softbank BB (Japan), SonaeCom/Clix (Portugal), Aksoran (Kazakhstan), and Woosh Wireless (New Zealand). Vendors entering the Alliance include Dovado (Sweden and United Arab Emirates), Ericsson (South Africa), Kathrein Werke (Germany), LG Electronics (Korea), Landasan (Malaysia) and Wireless System Integration (Sweden). They join founding members Andrew Corporation, Axcera, Fastcomm, InCode Telecom, IPWireless, MRiC, Possio, Samsung, and UTStarcom.

Despite this impressive support, UMTS TDD still has to prove itself in the mobile arena. All the deployments of TDD technology to date have been designed to deliver broadband wireless services to businesses and consumers in fixed locations. The operators involved are either incumbent fixed-line operators looking to rapidly expand their broadband coverage without digging up the roads, or new players looking to deliver high speed data quickly and economically to currently underserved areas. Major vendors such as IPWireless say they have carried out a number of trials with GSM/3G operators but so far no major mobile operator has publicly committed to TDD deployment in the near future.

Evolving WiMAX
WiMAX is the culmination of a number of previous variations of the IEEE802.16 standard. The first version, 802.16a, was intended to be an alternative to cable and DSL, capable of delivering higher capacity at lower cost using wireless. There were various revisions to the original standard which resolved issues such as testing and quality of service, ultimately leading to the WiMAX standard. The main proponents for WiMAX are Intel and Nokia, although latterly a number of operators have joined the WiMAX Forum and several are planning WiMAX trials. The early versions of WiMAX will work with fixed antennas, both outdoor and indoor, to deliver broadband wireless access to consumers and enterprises. The technology is also designed to provide backhaul for public WLAN installations. The third phase of development will be the 802.16e version which will support low mobility and roaming.

The WiMAX technology seems to offer nothing but benefits. Maximum data rates are claimed to be up to 100Mbits/s and the coverage could be up to 28 miles from the base station. The enhanced range in particular does raise some issues. To achieve a range of 28 miles will need much higher power outputs raising questions over interference and safety. Fairly high antennas heights will be required, bringing into play problems over costs and local objections. The final issue over WiMAX concerns the frequencies at which it is intended that the technology will operate. According to proponents of the technology WiMAX can operate in licensed bands between 2 and 6GHz or in unlicensed spectrum. All well and good, but unless this broad canvas is greatly reduced and refined, there will be long term issues over interference, device availability and cost.

The IEEE802.20 committee's aim is to develop and standardise a new technology that will deliver data rates of up to 1Mbits/s, with full mobility at vehicular speeds, a range of up to 15km, operating in licensed spectrum below 3.5GHz. The standard is to be specified from the outset to be fully IP. The technology has heavyweight supporters including Cisco, HP, Nextel and Flarion.

Leading the field?
Currently the favourite to win the air interface battle is OFDM. As already mentioned, Flarion's patented version of the OFDM technology – 'Flash OFDM – is already enjoying some success. In addition to the trial in Japan with Vodafone, US operator Nextel is carrying out a trial in North Carolina, USA in conjunction with Flarion which is claimed to provide 1.5Mbits/s download speeds. In Washington DC, Flarion and Motorola have built a high speed data network for what are called 'first responders' - that is fire, ambulance and police. The US Homeland Security Agency is looking at deploying such networks on a national basis in order to beef up the response to terrorist attacks.

When considering the future direction of their networks in terms of broadband wireless services, the choices and permutations available to operators seem endless. It is also unfortunate that decisions on the way forward will have to be made in the very near future - at a time when operators are already fully focused on the problems and costs involved in rolling out their WCDMA-FDD networks. With vendors on every side claiming that their solution is the only true way forward it is difficult not to feel a twinge of sympathy for the operators.

Difficult – but not impossible.
Ian Channing