What does it take to migrate from WiMAX to LTE?

Back in March I wrote a column on how TD-LTE was threatening to steal the spotlight from WiMAX by offering operators virtually the same technology and performance in a  more attractive wrapping--i.e., access to a larger market, more vendors and roaming opportunities. The article was mostly speculative and published before Yota's announcement to roll out LTE, and before Indian BWA operators disclosed their interest in TD-LTE. I was surprised to see how quickly in the past few month WiMAX operators have started to talk openly about the transition to TD-LTE or even to FDD LTE, and how they increasingly take for granted the ability to move seamlessly to LTE.

WiMAX and LTE share most of the RAN and core network, so the transition path to LTE will be smoother than, for instance, the transition path from HSPA to LTE. The trends towards RAN-agnostic base station platforms that can accommodate multiple wireless interfaces and towards user devices that support multiple interfaces further facilitate this transition. A few years ago, a RAN upgrade would have required the deployment of a separate overlay network that would have been difficult if not impossible to integrate with legacy networks.

Yet the transition to LTE will not be painless, and WiMAX operators should start planning for it now, so that they will be ready for it, if and when they decide to move to LTE. In the rest of this article I will go over some of the key decision points for WiMAX operators considering a transition to LTE.

Network coexistence: overlay or swap?

In the long term, most WiMAX operators that deploy LTE are likely to aim for a complete swap to LTE, simply because supporting two comparable networks adds unnecessary costs and complexity. In the short to medium term, however, operators may decide to build an overlay LTE network that coexists with their WiMAX network. This is the path chosen by Yota in Russia. The operator plans to deploy LTE in markets where it does not have WiMAX or as a second overlay network where WiMAX is available. This is a solution that guarantees a smooth transition for subscribers, who will keep using their WiMAX devices, and allow new subscribers to choose between WiMAX or LTE. However, only operators with sufficient spectrum can afford an overlay network. In addition, an overlay network requires a higher capex, as new base stations need to be installed, even if the deployed WiMAX base stations are upgradeable to WiMAX.

The alternative path is to plan for a network swap. Ahead of the swap, the operator deploys devices that support both WiMAX and LTE and then it gradually upgrades the base stations. Installation and equipment costs are lower, especially if the base stations can be upgraded and no additional spectrum is required, as long as TD-LTE is used or the operator can use the spectrum for FDD-LTE (see below). A swap requires more intensive planning, to ensure that subscriber devices all support both interfaces and will successfully switch to the new interface as it is turned on.

Spectrum: TDD or FDD?

WiMAX operators using the IEEE 802.16e version of the standard (Mobile WiMAX) use TDD, unpaired spectrum, with the same channel used for uplink and downlink transmission. The most common version of LTE, the one deployed by Verizon and by most mobile operators, uses FDD paired spectrum, with one channel used for uplink transmission and one for downlink transmission.

A second version of LTE, TD-LTE uses TDD spectrum and it will be widely deployed in China. Also, it is the version of LTE that Clearwire appears to be most interested in (but it may deploy FDD LTE as well). The transition path from WiMAX to TD-LTE is more straightforward. The same spectrum band can be used and infrastructure equipment upgrades are possible (although not necessarily available, depending on the vendor and on the version of the equipment).

Moving to FDD LTE may require the acquisition of new spectrum--this is the case for Yota--as regulation typically dictates whether spectrum bands can be used for TDD or FDD. Furthermore, acquiring new spectrum is difficult and expensive (FDD spectrum typically costs more than TDD spectrum).

The advantage of FDD LTE is that the market is considerably larger and equipment is already commercially available. This makes a transition in the short term possible, and widens the choice of devices and the revenue opportunity for roaming.

Base stations: what can be upgraded?

Despite the similarity in the underlying technology between WiMAX and LTE, not all WiMAX base stations can be upgraded to TD-LTE. Most of the early WiMAX base stations cannot be upgraded (or upgrades are partial, and require a substantial equipment cost), but many vendors now offer LTE-ready base stations. Upgradability to FDD LTE is considerably more difficult because of the use of two channels by this technology.

Core network integration

As both WiMAX and LTE use an all-IP core network, upgrades are possible, but may entail the replacement of some core elements. Integration among the WiMAX and LTE core networks is also possible and desirable in the case of overlay networks to enable subscribers to use both seamlessly. There are some differences in the core network architecture of WiMAX and LTE, however, and these require a clear transition plan and the selection of equipment that facilitates this process.

Devices: a roadmap to multimode?

WiMAX devices currently in the market do not support either TD-LTE or FDD LTE. WiMAX operators that wish to deploy an LTE network in the near future have to replace the devices if they swap networks or will not be able to enable roaming across networks for their subscribers who wish to keep their current devices. As WiMAX operators are not planning to swap networks in the short term, this is not a major concern. However operators considering an LTE overlay network should encourage subscribers to adopt multimode devices as they become available, so that by the time they are ready to launch the LTE network most subscribers have devices that can access it.

The increasing availability and affordability of multimode devices makes the transition to LTE more attractive to WiMAX operators, as operators are free to move to LTE at the pace they choose, taking advantage both of the wider choice of WiMAX devices today, and of wider availability and possibly lower cost of LTE devices in the future.

Timeline: when does it make sense to switch?

The timing of the move to LTE depends on multiple factors, such as spectrum availability, funding, or market conditions. These factors vary from country to country, and from operator to operator. For most WiMAX operators, it is too early to move to LTE, because the technology is still less mature than WiMAX, and its ecosystem is still evolving. Only once the technology is widely deployed and device selection is at least as good as WiMAX it will make sense for operators to move ahead with the transition.

The operators' choices will vary depending on requirements, funding, spectrum assets and overall market conditions. Many operators will not have many open options, mostly because of restrictions in spectrum availability. In all cases, however, WiMAX operators stand to benefit to start their assessment and planning processes now, to minimize the future cost and the disruption of the transition to LTE.

Monica Paolini, PhD, is the founder and president of Senza Fili Consulting and can be contacted at monica.paolini@senzafiliconsulting.com. Senza Fili Consulting provides expert advisory services on wireless data technologies and services.