To upgrade or to migrate: that's the LTE question

Many operators have said they "are committed to deploying LTE", but when pressed for details they often say they have no firm plans and that deployments will likely be sometime between 2013 and 2015.

HSPA+/HSPA Advanced has been considered for some time, especially since an HSPA+ network is backward-compatible with HSPA, allowing existing HSPA handsets to continue working. Also, base stations for all leading vendors are software upgradeable, resulting in lower costs and more seamless upgrades than a move to the LTE RAN.

LTE, however, makes all existing handsets unable to work. Users have to be either transitioned over the normal upgrade turnover time period for new multimode phones or provided with new phones in a more rapid swap out. If the operator opts for the swap out, they have to pay users to upgrade or face defections to other operators.

Long-term gains 

What makes this more complicated is that HSPA+ provides bandwidth that is close to that of LTE. So, why make a move to LTE (or Wimax) at all? Part of the reason is that these networks will become the long-term evolutionary path to even higher bandwidths.

However, many carriers do not need this for their mobile operations. If the operator also wants to supply bandwidth-intensive applications like enterprise data networking or if demand grows as fast as that for wired internet connections, they may find themselves in a battle with a competitor that does offer LTE or Wimax.

Similarly, LTE and Wimax might be pegged as similar to HSPA+ at 3.5G or perhaps a bit higher, but certainly not 4G. The decision to move to LTE is more complicated than the issue of bandwidth improvement over HSPA+ or of handset transition. Both LTE and Wimax are flat IP networks that incorporate quality of service and can provide very good VoIP service. A major issue for 2G-3G operators is how they maintain current revenue. If customers are transitioned to VoIP they will likely expect it to provide unlimited calling similar to internet broadband-based VoIP service plans.

Several factors weigh into decisions about what network makes sense for new deployments versus upgrades. How the HSPA+ or LTE migration issue is framed makes a difference. From a more technical perspective, it can be said that LTE does not yet have the proven capability to handle comparable voice calls. However, all studies and examples of early deployments show that voice capacity will be higher and better quality than for 2G-3G. Like every network that has ever been deployed, LTE has to be developed, debugged, deployed, tuned and refined so that coverage is comparable to 3G and adequate QoS is maintained. If the appropriate funds and efforts are expended, there is no roadblock in terms of the technology.

Some of the reluctance and concerns about problems are likely to stem from resistance to change. It is good to look back just a couple years ago and compare what major suppliers and operators were then saying about the prospects for Wimax and LTE technology - many said that these were inferior or had no advantages over HSPA or EVDO.

Since then, AT&T and Verizon have acquired spectrum at a premium cost to deploy LTE while other operators have announced commitments. The tone has definitely changed although nothing much has in terms of the technology of these systems, except the natural process of development.

AT&T is a prime example of this process. Its director of networks recently said that although he considers LTE as a more capable technology platform, the company will still go ahead with enhancements to its 3G network.

The reason for going ahead with upgrades rather than converting to LTE is that the move to LTE would be disruptive to current handset users and costly for AT&T. Robert Syputa is senior analyst and partner at Maravedis