At the 3GSM conference two weeks ago, there was a lot of effort expended by the 3G community to convince us that HSDPA and its successors have caught up with the mobile WIMAX roadmap, making it equally valid, and heavily backed, as the basis of true personal broadband. While indoor penetration remains an issue, and HSxPA will not address the uplink until 2007, there has certainly been an acceleration of the evolution of W-CDMA in response to disappointing results from 3G build-outs, and the threat of WiMAX (see Wireless Watch February 22 2006 issue). A week later, at the WiMAX Summit in Paris, all the focus was on the advantages that 802.16e has over its 3GPP counterparts, in terms of fixed broadband performance, early adoption of smart antenna techniques, spectral efficiency and rural coverage. Most of the points made were true, and did not disguise the fact that the cellular world has galvanized itself into action to retain control of its customer base, and that these are now two very strong technologies, with different strengths and weaknesses, but many of the same medium term directions (notably, towards greater fixed/mobile convergence and the adoption of key technologies such as OFDM). Increasingly, then, there is a case for looking to converge these would-be 4G platforms, one driven in particular by the major multi-network vendors such as Motorola and Siemens, which will thrive on integrating various systems to provide a comprehensive broadband architecture.
A single roadmap would not necessarily involve a single technology in the short term, but a set of key elements to which different networks would conform, supporting roaming, hand-off and integration at the core (as is already emerging in the Wi-Fi/WiMAX and Wi-Fi/cellular markets). Companies like Cisco and Siemens are already active in bringing WiMAX and 3G together in real deployments, and it seems only a small step until this happens in the standards bodies too, especially as the suppliers look to reduce their R&D budgets and avoid duplication of effort.
Ron Resnick of Intel, president of the WiMAX Forum, told reporters on the Wireless Broadband Analyst that a Forum member recently proposed a merger of the 802.16e efforts with those of the 3GPP LTE (Long Term Evolution) group, which is devising a ‘3.9G’ platform using many of the elements also associated with WiMAX, such as an OFDM physical layer in the downlink and MIMO smart antennas. A key contributor is Nortel, which also has high interest in WiMAX and is reusing some of the R&D it undertook for 802.16e in its HSOPA architecture, which has been adopted for LTE.
Despite these increasing levels of synergy, there will be many political obstacles to a merging of the two mobile broadband future standards, not least the desire of the main beneficiaries of cellular network investment, such as Ericsson, to keep the market tightly under their control. Operators could be wary too, especially those such as NTT DoCoMo which contribute heavily to 3G+ research. A converged platform would be likely to span far more spectrum options than LTE and so open up opportunities for new competitors from the wireline, start-up and broadcast sectors.
Resnick said that, while the Forum believes in the eventual convergence of LTE and 802.16e, to achieve lower prices, higher volumes and interoperable devices, the 3GPP had so far not looked favourably on the idea. WiMAX opponents would say this was because the 3GPP technologies are so far ahead in terms of market share and future evolution that they have no need to accommodate an alternative; WiMAX backers that they do not wish to share their market with a superior platform supporting new, largely PC/IP-oriented vendors and business models, and one that is two years ahead of LTE.
The reality is that these two technologies look increasingly similar, and many vendors are involved with both, but the key stumbling block may be the need for LTE to be backwards compatible with W-CDMA, a necessity that would constrain new WiMAX systems and operators. There are technical difficulties in the way of full interoperability, namely the likelihood that LTE will use OFDMA only in the downlink, preferring an uplink technology called SC-FDMA (single carrier frequency division multiple access), which it claims puts less strain on the battery.
But the real obstacles are political, not technical, and eventually the pressure from carriers will push vendors towards merging their efforts. Operators of all kinds will want to be able to mix and match different networks, in order to deliver a range of fixed and mobile services for different and increasingly narrowly defined customer bases. This is already seen in an embryonic form in the combinations of Wi-Fi, 3G and DSL in projects such as BT Fusion, but far greater flexibility will be required in future to keep operators competitive. They will push for a set of networks that delivers everything ‘ true broadband, gigabit fixed indoor coverage, high speed mobility, high definition television, security and quality of service, low cost per bit, and so on. One technology will almost certainly compromise in one area or another, while a seamless combination looks like the way to go. In the short term, the WiMAX Forum’s apparent enthusiasm for convergence may suggest it is feeling threatened by the accelerated evolution of the W-CDMA platform. In the longer term, though, WiMAX is heavily supported and has many real strengths that will take their place in the next generation network. The speed with which the 3GPP wakes up to the need for cooperation will largely depend on the real world performance of 802.16e once systems are delivered. If it looks close to HSDPA in operator benefits, we can expect a two-headed race to persist for some time, as the technologies will be largely in sync in terms of roll-out. If it looks closer to LTE ‘ and much of the extensive investment in 802.16e enhancement being undertaken by major vendors this year is aimed in this direction ‘ WiMAX will have taken a step ahead of the 3GPP, with LTE unlikely to appear in systems for at least two years (except in the interim form of TD-CDMA release 7). At this point, we would expect the drive for convergence to gather real momentum, with a concerted effort to sideline the Qualcomm technologies being seen by many in the 3GPP camp as an added bonus.