Will you be getting a 3G handset for Xmas?

Citation

Curwen, P. (2005), "Will you be getting a 3G handset for Xmas?", info, Vol. 7 No. 2. https://doi.org/10.1108/info.2005.27207bab.001

Will you be getting a 3G handset for Xmas?

Will you be getting a 3G handset for Xmas?

So long has passed since the first 3G licences were issued, and so many deadlines have passed, that most readers may not have noticed that 3G is in a meaningful sense finally among us. It is safe to assume that large numbers of people have received 3G handsets in their Xmas stockings, and deciphering the accompanying handbook should have kept them happily amused over the festive season – unless, of course, the proud owner was a teenager – in which case, it would all be intuitively obvious.

Before moving on to examine the data as set out in Tables I and II, some comments are in order. First, it must be borne in mind that the roll-out of cdma2000 1xEV-DO is merely an upgrade on a (mostly) pre-existing cdma2000 1xRTT network. cdma2000 1xRTT is best regarded as a version of 2.5G roughly comparable either to GPRS or EDGE according to preference, but neither of the latter can be upgraded to W-CDMA which operates over a different spectrum band. This would suggest that the tables should contain far more entries for cdma2000 1xEV-DO than for W-CDMA (or UMTS as it is generally known, especially in Europe), especially since it is not normally the case that a new licence is required to upgrade a network whereas it is invariably required to operate over a new spectrum band. However, account must be taken of the fact that GSM, in which W-CDMA is rooted, is by far the predominant 2G technology worldwide and that it is used exclusively in Western Europe where the process of 3G licensing largely originated. In addition, W-CDMA is the chosen technology of two of the three Japanese incumbents and was licensed in South Korea at an early date. cdma2000, in contrast, is largely to be found in the USA, where the 3G licensing process has been, to put it politely, a mess (see the article by this author in Vol. 7, No. 2), and in Latin America where mobile networks are for now relatively unsophisticated.

Table I W-CDMA launches

Second, it is impossible to produce a definitive list of 3G launches. Hence, although a very small number of alternative lists are available in the public domain, it is absolutely crucial to read their footnotes as it is not so much that the lists are incompatible, either among themselves or with Tables I and II, but rather that they are based on somewhat different definitions. In Tables I and II, the dates apply wherever possible to what is commonly referred to by operators as a “commercial” launch. However, this may, in practice, be something much closer to the concept of “pre-commercial” and, in any event, generally applies to a (sometimes quite small) section of a complete network – the minimum population coverage has never been standardised even in Western Europe. By and large, something akin to a full national roll-out will take between one and two years from the official launch date.

Table II cdma2000 1xEV-DO launches

Another contentious issue is the speed of data transfer. Not unnaturally, operators are keen to launch their networks once they are technically operational, and this often means that their typical data speed is only 128 kbps. Strictly speaking, that is 2.5G, and a minimum of 384 kbps is required for 3G. In any event, there is generally some difference between a claimed speed and the somewhat lower average speed enjoyed by users when more than one of them is logged on. This issue is relevant both where handsets are in use and for data cards in laptop computers, but the launch date controversy is also clouded by the preference of operators to launch initially into the enterprise market with data cards, to be followed many months later with a “commercial” launch involving handsets.

Hopefully, the footnotes to Table I clarify the basis on which entries have been made, and on the whole the table tends to contain a few more entries (in respect of W-CDMA) than that produced by, for example, www.gsacom.com, which defines a “commercial” launch as one meeting the following criteria:

1. 1.

   Anyone can subscribe to the service (as against a trial).

2. 2.

   W-CDMA handsets or terminals are available to users.

3. 3.

   W-CDMA handsets are sold rather than rented to users.

4. 4.

   The operator has made a public announcement.

5. 5.

   The service is charged for.

As can be seen, there are 75 entries in Table I which includes information up to 10 December 2004. It is worth noting that a further 14 launches had been announced for 2004, although very few, if, any are now likely to take place. In addition, 20 launches have already been announced as definite for 2005. This means that by the end of 2005 the number of W-CDMA launches will be a minimum of 110 but quite probably rather nearer to 130. Although most of the launches so far took place during 2004, there are those who will quibble about whether W-CDMA has as yet truly arrived. Within six months there will no longer be room for argument.

Table II confirms that there have been few cdma2000 1xEV-DO launches so far. A further four had been announced for 2004 (Telstra, Bermuda Digitel, Sprint and Movilnet of Venezuela) and only one definite for 2005 (Vodafone in New Zealand). Hence, despite the constant barrage of publicity about the relative ease of upgrading cdma2000 1xRTT to 1xEV-DO, it is clear that the great majority of CDMA network operators are to be found in countries where the market for high-speed data services is limited, and hence there is no economic case – at least, for the time being – for a speedy roll-out of 1xEV-DO.

We must conclude that in the 3G stakes there is only one clear winner – W-CDMA. This may not accord with the impression gained by casual readers of the literature on 3G, but there is a lot of obfuscation deriving from the issue as to the dividing line between 2.5G and 3G. GPRS is on the slow side compared to cdma2000 1xRTT, but the latter certainly does not qualify in this author’s eyes as a 3G technology. Furthermore, EDGE has now entered the equation in a significant way as, in general, a half-way house to full 3G for GSM operators – 2.75G if one must – that requires neither new spectrum nor new licences. According to the author’s database, there have been 45 launches of EDGE to press (www.gsacom.com says 38), but a further 40 had been announced for 2004 so the number is probably already rather higher and will reach 100 early in 2005, rapidly overtaking cdma2000 1xRTT. At the end of the day, the benefits of inter-network roaming provide a massive incentive to roll out EDGE and W-CDMA, so the above developments should not really come as all that much of a surprise.

Peter CurwenVisiting Professor of Telecommunications, Strathclyde Business School, Glasgow, UK. E-mail pjcurwen@hotmail.com