After all of the buzz over the past couple of years, at the end of 2009 TeliaSonera launched the world's first commercial LTE network in Oslo, Norway and Stockholm, Sweden. Four more operators are lining up to launch their own LTE services this year. So what will the mobile data scene look like in a couple of years, asks Scott Siegler, Senior Analyst, Mobility Infrastructure, Dell'Oro Group.
There is little doubt about the public's hunger for mobile data services – the traffic continues to double year on year and is forecast to reach nearly 4 billion gigabits per month by 2014 – and LTE promises peak data rates exceeding 100Mbps, as compared to the best current HSPA+ peak rate of 42 Mbps.
Does that mean it's ‘all systems go' for the new technology? That no self-respecting mobile carrier will be without an LTE service for fear of being left behind in the competition for customers? The answer is no – and any claim that LTE will be sweeping all before it in two or three years time is simply setting the stage for LTE being seen as a damp squib that failed to live up to expectations.
Don't be carried away by the excitement of the new technology and its potential. There are several challenges still to be addressed, and beyond that there are sound business reasons for holding back on LTE deployment, as I will explain. There is no doubt that LTE is the future for mobile telephony, but it is better to think in terms of a slow burning fuse for the next couple of years.
For a start, we face the classic chicken and egg situation: who wants to invest in launching a costly new service when there are so few devices on the market to utilize it? And who wants to invest in making LTE-ready user equipment when there are so few services available? It is one thing to launch a few pilot services, but quite another to allow time for the whole ecosystem – including the radio access network, the packet core network and the LTE-enabled handsets and devices – to be developed, rolled out and to reach critical mass. Many operators are delaying LTE until a diverse set of devices is available to attract subscribers. The only LTE devices currently in the pipeline are data cards and dongles, with LTE handsets not expected until 2011 at the earliest, and more likely in 2012 on account of the need to pack
greater complexity and multiple antennas into a small enough package, as well as the need to standardize voice and SMS support over LTE. But this problem of launching any new service is not new – similar issues slowed the initial uptake of 3G services.
Another reason for delay is the current lack of spectrum availability. This applies especially in Western Europe where the ‘digital dividend' spectrum will not be available for use until 2012. Some countries in Western Europe have already licensed the higher 2.6 GHz spectrum, but the majority have still to auction it – and in any case the lower propagation properties of RF signals in the higher spectrum bands require more base stations to provide the same coverage as the lower bands.
So there are technological challenges still to be addressed, but there is also a strong business case for providers not rushing prematurely into LTE. On the one hand there is the prestige attached to launching an LTE service ahead of the game, on the other hand there is the pragmatic short-term option of maximizing the ROI in existing infrastructure – and the majority of service providers will be taking the latter course.
Consider the fact that the majority of operators currently planning to deploy LTE already have robust and extensive HSPA networks in place. Purchasing new spectrum in addition to brand new infrastructure will cost billions of dollars and the operator needs to evaluate over what time period the hefty up-front costs will be recouped by their LTE services.
While some operators do have plans to migrate their networks from HSPA 7.2 directly to LTE, most plan rather to migrate to HSPA+ as an interim step. With only a software upgrade, HSPA+ can support peak downlink rates of 21 Mbps, using only a 5 MHz channel in spectrum the operators already own. With slightly more incremental cost, operators can upgrade their HSPA+ networks to support peak downlink rates of 42 Mbps (it is actually possible to support twice that rate, but I don't see this option being taken up on a big scale before LTE arrives in force).
I base my case on two recent Dell'Oro Group reports: Five Year Mobile Forecast Summary January 2010 and May's Mobility Infrastructure 1Q10. As the figure shows** the report anticipates a steady return to overall growth in the infrastructure market following the steep decline during 2009.
While forecast levels of GSM and CDMA spending have been significantly reduced in this latest report, WCDMA revenue is expected to grow from $15 B in 2009 to $32 B in 2014. Despite the arrival of LTE, there is no evidence of providers choosing to restrain current investment in anticipation of tomorrow's 4G technology. On the contrary, the report suggests that the heavy 3G spending will be the dominant driver of the market for the foreseeable future.
As a result, although the report sees the first signs of LTE revenue showing in 2010 – driven by early adopters like Verizon – the real growth inflection point does not show until 2012. That is when the digital dividend spectrum first becomes available in Europe, and when the report expects the majority of operators in Western Europe in addition to China Mobile to begin LTE roll-outs. By 2013 all three Chinese operators are committed to deploying LTE and China is expected to become a significant driver judging by the impact their 3G deployment had on worldwide sales.
So, my advice would be to forget the aggressive timelines being proposed for deploying initial LTE networks in 2010, the findings in these two reports explain why – as well as offering a deeper understanding of the providers' situation.
Unlike with the 3G market which is driven by the explosion in the demand for mobile data, it will be coverage rather than capacity that drives initial LTE deployments.