The more similar the next generation wireless platforms become, the more important politics and spin become in differentiating them. So in the week of WiMAX World, Qualcomm and Ericsson sought to divert some attention from Mobile WiMAX towards their own favourite OFDM-based systems, UMB and LTE respectively. The CDMA Development Group, with surely non-accidental timing, published its official specification for UMB, while Ericsson did much to maximize the impact of comments made the previous week by Verizon and Vodafone, indicating that the world’s most fervent CDMA supporter, Verizon Wireless, would was likely to go with LTE for its next iteration. While the Verizon Wireless hints were certainly not as definitive as some reports would have us believe, they do represent a major shift of attitude from the US’ second carrier, which will only pile more pressure on Qualcomm in a week where, with typical schizophrenia, the chipmaker raised its forecasts for the coming quarter, while facing intensified antitrust probes from the European Union. However, Verizon Wireless is clearly far from clear technical decisions, and the significance of its comments is to reinforce the whiphand that the large cellcos currently have, playing off one vendor faction against the other in order to increase their own influence over the next generation network. They will be determined to use this power more effectively than they have in the past to ensure that suppliers work to their specifications.
In the 3G generation, most cellcos were stuck in one technology path from which it would be difficult and expensive to veer away; with the pre-4G options looking so similar, the difference in the transition from 3G to any of the three main options is considerably narrowed. Therefore, the operators can keep their choices open and make maximum capital out of the uncertainty of the supplier community. A few weeks ago it was Vodafone that threw a cat among the pigeons by indicating that it would be prepared to change the habits of a lifetime and adopt a multi-technology strategy in the future, potentially introducing WiMAX as well as LTE to the mix. Then it was the turn of Vodafone’s US joint venture Verizon Wireless, which told a Goldman Sachs investor conference that both sides of the partnership were involved in mutual LTE development work. Vodafone CEO Arun Sarin and his Verizon counterpart Ivan Seidenberg said that the companies were committed to having a common platform for the next generation, and hinted strongly that this would be LTE.
Many trials and technology choices lie between this very general statement of intent and real contracts, but the statements are undoubtedly a huge blow for the hopes of the CDMA community that its platform will last into the fourth generation. Of the CDMA majors around the world, Sprint Nextel has already chosen WiMAX for its ‘4G’ system’; KDDI will pursue a multi-network strategy if its regulator allows; the Koreans are also putting significant effort into WiMAX, as a system in which they have major IPR advantages; and the future of CDMA regulations in India remains in turmoil. Some major operators, such as Vivo in Brazil and Reliance in India, have already indicated plans to defect. All this has left Verizon and, to a lesser extent, Bell Canada, holding many of the cards in deciding whether there will be sufficient market to justify investment in commercializing UMB. Without a major carrier committing to the technology, the primary vendors, such as Alcatel-Lucent and Nortel, will not commit either, and Qualcomm will be forced to focus on attaining leadership in markets where it has less direct control, notably LTE.
For Sarin and Seidenberg, the main motivation behind the carefully timed statement of harmony was to quiet restive investors, especially on the UK side of the Atlantic, and quell uncertainty over whether Vodafone would sell its 45% stake, something Verizon has often pushed for in the past. Now it seems the two partners are determined to make a go of their union. Asked if the complementary network evolution indicated a long term relationship between the two companies, Seidenberg said that was ‘fair’ and that Verizon was looking for stability in its relationship with Vodafone.
‘Going through a common platform is nothing more than the industry realizing that we can stimulate expansive growth by having a common platform and having the best networks,’ he added.
Although the LTE co-development gained prime billing, the two cellcos will also test WiMAX and other ways to integrate their future systems. Seidenberg told the conference: ‘I think we’re going with some trials with them on WiMAX. We’re doing some other trials with them on selling into the enterprise market. We have rallied around this new 8830 BlackBerry that is the global BlackBerry for us [the 8830, which supports GSM/CDMA roaming]. So we have very good operating relationships with them right now.’
Verizon Wireless and Vodafone can achieve significant economies of scale by adopting one platform. The former has 62m subscribers, while the latter has more than 200m round the world. Sarin said: “It makes complete sense for us to go from HSDPA to LTE and it makes sense for Verizon to go from EV-DO to LTE’, although neither CEO expects mainstream deployment for 3-4 years at least. This timescale could be accelerated if Verizon Wireless gains large swathes of spectrum in the 700MHz auction planned for January, since it will be under FCC deadlines to deploy networks, and would likely look to implement a pre-4G system right from the start.
Verizon Wireless has never before given any real indication that it might stray from the CDMA path towards WiMAX or LTE, although with doubts over the future commercial viability of UMB, it has hinted it could be deterred by the threat of lower volumes and therefore higher prices in this sector. But there are clearly many practical decisions and much internal politics between a CEO statement to investors and real deployments, and after the Goldman Sachs event, Verizon Wireless was already distancing itself a little from the initial comments, with spokespeople stressing that no specific technology has been selected. This ambivalence will be important to maximize Verizon’s bargaining power – as the largest CDMA customer and influencer, it has the potential benefit of keeping a technology over which it has huge control. In general, the venture has been rather more conservative about major technology upgrades than its rivals, and only a few months ago, the cellco stressed that it did not believe it would need a network beyond CDMA EV-DO Rev A – which it is currently deploying in many US markets – for many years (see Wireless Watch May 14 2007). Dick Lynch, then CTO of Verizon Wireless, and now CTO of the whole group as it moves to convergence, said in an interview then: ‘Let’s ask ourselves what do we need this technology for? What do our customers need that they can’t get on 3G? I think the jury is still out there.’ This was in clear contrast to the aggressive ‘4G’ roll-out plans of its main CDMA rival, Sprint Nextel. In August, Lynch marked his new role with similar comments, telling an RBC Capital Markets conference that WiMAX was ‘just another air interface technology. WiMAX, LTE and UMB are essentially 4G wannabee technologies. They all use OFDM technology and perform the same way in the same amount of spectrum. Would I use WiMAX with Verizon? Possibly. If I had TDD spectrum.’
Only a few weeks after that, at the NXTComm show, Seidenberg was sounding more gung-ho about 4G, saying: “We’ll be ready for 4G by the end of the decade.”
This clearly indicates the gulf between the political clout that CEOs can gain from public comments, and the realities faced by executives like CTO Lynch. Beyond the manoeuvrings of the joint venture itself though, there is a broader agenda that all major cellcos have close to their hearts and in which Vodafone/Verizon Wireless aim to be direction setters. This is to harmonize standards, in order to reduce cost and risk, and to ensure the major carriers, not the major vendors and IPR holders, have the dominant influence over the shape of those standards. Hence, the main vehicle for this effort, the operator-dominated NGMN (Next Generation Mobile Network) forum, is potentially more important than all the workings of the standards bodies as 4G starts to be defined. Verizon Wireless has been on the harmonization track since the 1990s, when it called for unified 3G standards, an effort scuppered by vendor politics and technology differences between carriers.
All is not lost for UMB, but it will require a superhuman effort to make it a major force in 4G. Its main potential weapons are its time to market potential and its spectral efficiency for operators with narrow bands of spectrum. The former will be critical for carriers like Verizon Wireless if they aim to leapfrog rivals by getting next generation services to market first ‘ an ambition where, in the US, it has already been outdone by Sprint ‘ or if they are under regulatory mandate to deploy quickly. UMB is farther ahead than LTE in terms of real specifications, and is likely to have a more streamlined roadmap because of the smaller community and the coordinating hand of Qualcomm. WiMAX, of course, is further ahead still and will use its status as the first ‘real’ pre-4G network to argue that UMB is redundant, leaving the Qualcomm camp under pressure to put demonstrable clear water between its technology and 802.16e in terms of price/performance and efficiency, and ease of migration from previous EV-DO releases.
Though real systems will be necessary to for this, at least formal specifications are now available. UMB combines OFDM on the downlink with MC-CDMA on the uplink. Last week, the main bodies supporting it ‘ the CDMA Development Group (CDG) and the 3GPP2, published the Ultra Mobile Broadband air interface specification. This claims that UMB is the first IP-based mobile broadband standard to enable peak download data rates of 288Mbps in a 20MHz channel, and with seamless integration with CDMA2000 and EV-DO.
Among the key specs are peak download speeds of 288Mbps and upload of 75Mbps; delivery of high capacity voice and data at over 300 kilometers per hour; low latency of 14.3ms average over the air with reduced jitter; supportfor up to 1,000 simultaneous VoIP users within a single sector.
Of course, few lack confidence in Qualcomm’s ability to deliver effective technology for UMB, as for other platforms. Where confidence in the company, and therefore the systems with which it is most associated, is shaky is on the IPR side, where regulatory pressure and legal challenges are piling up against the chip giant’s business model. The latest is an expected move from the European Union, which has been petitioned by various vendors, led by Nokia, to investigate alleged antitrust at Qualcomm. Today the European Commission said it had elevated its investigation to ‘priority status’, making decisions early next year a likelihood. The complaints, brought by Nokia, Ericsson, Broadcom, Panasonic, NEC and Texas Instruments, center on allegedly excessive royalty levels, and breaking of rules on ‘reasonable and non-discriminatory’ licensing conventions.
Qualcomm raises guidance:
Every time a new blow falls on Qualcomm, the chipmaker manages to bounce back in terms of its current performance, even as shadows loom over its future structure. Last week it raised its guidance for the fourth quarter and full year, a move that is becoming almost routine. It now expects to make revenues at or above its previous guidance of $2.15bn-$2.25bn in Q4, on shipments of 67m-68m units of its MSM chips. A year earlier the company made revenue of $2bn. For the full year, it expects revenues of $8.72bn-$8.82bn, up from $7.53bn last year. The fact that 3G is the main growth driver means margins are expected by most analysts to have improved. Qualcomm noted that it recorded no royalties from Nokia for the fourth quarter, because of the two companies’ failure to resolve new licensing terms following the expiration of their agreement in April.
Qualcomm also hired a new top lawyer, Donald Rosenberg, who joins from Apple, where he has been general counsel for just 10 months. Qualcomm’s former general counsel, Lou Lupin, resigned in August for ‘personal reasons’ after a string of legal defeats in the chip company’s various patent related legal battles.