3G momentum has been a consistent theme in this column. However, as we have pointed out, strong uptake of 3G handsets is not always accompanied by sustained uptake of 3G services. It seems High-Speed Downlink Packet Access (HSDPA) technology may be an exception - launches in Europe have already aroused strong interest. Looking at the opportunities in Asia, we ask, is this (finally) 3G for real‾
HSDPA has come a long way, fast. Back in 2002-2003 deployment seemed a long way off, and we detected little sense of urgency in the 3GPP. Its members were occupied with dealing with the complexity of deploying W-CDMA. Two things happened to speed development of HSDPA: W-CDMA emerged from its teething phase and WiMAX came along. This latter technology, heralded by all as the 'barbarian at the gate' of the 3G citadel, was to finally eat the 3G lunch barely nibbled at by WiFi.
Fast forward three years to October 2006, and this scenario has changed considerably. There have been 52 commercial launches of HSDPA technology in 35 countries on three continents. And 117 operators have made HSDPA commitments in 54 countries. By contrast, mobile WiMAX operators have yet to see certified equipment.
Until now, most of these deployments have been in Europe: 26 out of 38 commercial services by end June 2006. There are two reasons for this. First, European operators have seen increasing interest in mobile access delivered to the laptop from their enterprise customers. Second, the complexity of building these technologies into the mobile terminal has meant that HSDPA datacards have been far more common than handsets, which have only started to appear in volume in the second half of the year. Consequently, operators have been building out the network with a view to corporate users - covering the transport corridors and airports and metro centers first.
In Asia, we expect to see a more heterogeneous deployment. Not all operators have the same ready installed base of corporate users. Many are focused on the consumer market. Yet HSDPA in the consumer market is a less obvious play. For some markets, it is deemed essential.
In Japan NTT DoCoMo needs HSDPA to offer full-track music download to compete with KDDI's hugely popular 'ez-chaku-uta full' music service. In South Korea SKT's capex guidance indicates that HSDPA is a greater priority than the locally developed WiBro service it is also deploying. And KTF sees the technology as a significant opportunity to compete with SKT on a level spectral playing field, as both operators will deploy at 2100 MHz, as opposed to the prime 800-MHz spectrum that SKT uses for CDMA services.
So far, so good. But we believe that the performance of HSDPA could create a few problems. Not because of service shortfall - Cingular is recording average end-user (not peak) downlink speeds of 400-700 kbps, bursting up to 1 Mbps. This is impressive - 3G for real. Yet these speeds mean that users will generate significantly more traffic than users on W-CDMA and GPRS networks.
We can already see the implications in the tariff structure of European HSDPA services. In Austria, ONE has a massive 10-gigabyte usage cap on its Mobile Breitband HSDPA flat-rate tariff, which is sold at 75 euros. MobilKom offers a 3.6-gigabyte cap on its tariff, available for 99 euros.
With HSDPA speeds set to increase through implementation of different category classes to a (theoretical) maximum of 14.4 Mbps, the extra traffic could have a huge impact on capacity requirements both in the radio access network and in the backhaul.