Devices and spectrum 2 thorns in the side of LTE

Devices and spectrum 2 thorns in the side of LTE

The Delta Perspective
The advent of LTE promises a quantum leap in mobile data speeds.  However, what the tagline does not state is that this increase in speed is accompanied by an increase in the speed of battery depletion, as higher speed requires more energy.  As LTE is making the news around the world, with more operators announcing LTE network rollouts, the main headache relates to the devices that support the technology and especially in relation to the choice of spectrum in which LTE gets deployed.
LTE is supposed to deliver superior throughput, relieve saturated 3G networks and decrease the unit cost of data.  Despite the numerous rollout announcements, the technology remains in its infancy and suffers from a lack of standard spectrum allocation, unlike 2G and 3G.  The reality is that high quality spectrum (<1GHz, which has superior propagation, both indoor and outdoor, hence reducing CAPEX requirements) is scarce and even sub 2GHz is relatively limited.  Consequently, LTE has to deal with the leftovers.  The main implication is the large spread of deployment frequency from 700 to 2600MHz.  As an illustration of the issue, the 3GPP (the LTE standards setting body) has allowed no less than 40 different bands around the world.

Regulators and operators are now following 3 broad spectrum strategies:
  • Refarming existing mobile spectrum: e.g. 1800 MHz in Europe, 1900 MHz in the US
  • Refarming non-mobile spectrum (mostly TV digital dividend): e.g. 700 MHz in the US, 800 MHz in Europe, 2300 MHz in SA (DECT backhaul)
  • Opening new spectrum: e.g. 2600 MHz, or 2300 MHz in China, India, Japan
The major consequence is that there is no global standard for LTE deployment globally or even within a specific country.  In the US for instance, while AT&T and Verizon are implementing LTE on 700 MHz, Sprint is refarming 1900 MHz and potentially Clearwire WiMax spectrum (2.5GHz).
This has, of course, material impact on the devices in terms of which frequency to support.  The handset OEMs need to balance the number of bands available with the battery life of the device.  The more bands the device is forced to accommodate, the more battery power is consumed, thus a  shorter battery life.  Whilst chipsets are available to handle GSM, UMTS, TD-LTE, FDD-LTE in multiple bands, to optimize battery life, the OEMs are selecting individual LTE bands that align to their primary target market.  For instance, LTE on the iPad 3 only works in the US.  The iPad mini also reflects the primary US target with LTE being available on Sprint’s 1900 Mhz.
As the technology matures, chipset efficiency will increase and more bands will become available.  The exhibit below illustrates that by end of 2014, some high end phones capable of accessing LTE globally will become available.

In the short term, OEMs are left with two options:
  • Regionalization of handsets:
The OEM creates different versions of the each model that play in different bands.  For instance the iPhone 5 comes in three flavours: US GSM, US CDMA, Europe/ Asia GSM.
This approach creates a lot of frustration. Apple, for instance, received a court order to remove the 4G label on the iPad 3 in Australia since the device was not compatible with the LTE spectrum available in that country.  Device manufacturers will have to build this capability cost- effectively as it increases logistic costs and presents customer satisfaction issues, especially in the high-end, frequent traveler segment.
  • A single chipset capable of multiple bands:
A chipset capable of multiple bands is already available but is impractically energy-hungry.  The idea here would be to create radio chipsets that are able to turn different bands on or off either manually (at first) and eventually automatically by communicating with the network or based on location provided by the operator or GPS.  This should unlock significant power savings.  Additional power-saving features are being tested as the STPTIC from STMicroelectronics, can electrically adjust the amount of energy transferred from the handset’s amplifier to the antenna to help maximize call performance when using multiple frequency bands.
It is clear that the first option is currently gaining the most momentum in the short term. However, given the proliferation of LTE bands (already ~40 bands; excluding the 18 3G bands already in use) and the scarcity of spectrum, LTE handsets will need to become a lot smarter in energy consumption.
As far as the operators are concerned, devices remain a critical part of the mobile ecosystem as depicted in the recent years with device successes such as the iPhone.  Considering that LTE is still in its early stage, some prudence is required regarding spectrum strategy.
The reality is that the large operator in dominant markets will drive the economies of scale on their own.  This means that large OEMs will likely follow them in the frequency deployment, as illustrated by AT&T and Verizon in the US, NTT Docomo in Japan and China Mobile. This becomes more delicate for third/ fourth entrant operators in large markets or operators in smaller markets.  Their relative scale will not be sufficient to drive OEMs to cater specifically for their chosen LTE spectrum.  Therefore, it becomes critical to adopt one of the following strategies:
  • Define an agreed frequency standard at region level (this is the path followed by Europe)
  • Lobby the Government/ regulator to expedite release of LTE frequencies that are in line with the major deployments in the region
  • Wait and see where the standards are emerging, at the risk of losing some market share, in the interim
Should these strategies prove unfeasible, the operators are likely to end-up with device portfolio disadvantages or paying handset subsidies to level the scale disadvantage, at least until the handset OEMs figure out how to crack the device battery life conundrum.