Carrier Aggregation over Unlicensed Spectrum

Posted on June 1, 2014

Recently I’ve been following the proposals to use unlicensed (i.e. Wi-Fi) spectrum to augment cellular capacity. The most prevalent camp proposes using LTE-Advanced carrier aggregation, introduced in 3GPP release 10, where the 5Ghz Wi-Fi bands are used for downlink data, and all other traffic, such as signaling, is ‘anchored’ on licensed spectrum.

The 5Ghz unlicensed bands offer close to 500Mhz of bandwidth, and in the U.S., South Korea, and China can be freely used without coordination with radar and safety services. This means that unlicensed LTE, or “uLTEA” as it’s being called, could be rolled out without having to make complex changes to the LTE specification in these countries. Other countries will have to wait for 3GPP release 13, which will handle the ‘listen before transmit’ requirements present outside of the U.S., South Korea, and China.

The two main advantages of this approach are:

  • It’s all LTE and fully integrated into the packet core, without any gateways or other complex coordination.
  • QoS is maintained by the licensed spectrum anchor point, so if unlicensed spectrum degrades, traffic and signaling quality can still be maintained.

This sounds great, but only really works if both the LTE eNodeB licensed and unlicensed radios are co-located in a small cell. Therefore, this whole strategy relies on the continued successful deployment of small cells. Also the small cell vendors that are going to provide uLTEA solutions need to get coexistence with traditional Wi-Fi figured out and rock-solid, otherwise the massively deployed base of Wi-Fi will be affected adversely with the loss of the 5Ghz bands.

Perhaps a minor technical point, but uLTEA is going to require radios in the handset that operate simultaneously on two (likely) quite different frequency bands, which could adversely affect battery life. Of course, OEMs have a track record of solving these issues.

So, where does this leave existing carrier Wi-Fi solutions, if a fully integrated and coordinated solution has all these advantages?

It’s clear that for mobile network operators contemplating the deployment of small cells, there would be no need for Hotspot 2.0, or indeed Wi-Fi. Additionally, roaming data services, a centerpiece of Hotspot 2.0, would be handled seamlessly, just like any other LTE roaming scenario. So, there is a chance that Hotspot 2.0 would be marginalized by uLTEA.

Adoption of uLTEA and a slowdown of Hotspot 2.0 will make traditional offload propositions from non-cellular provides such as Wi-Fi pure-plays and cable operators even harder to implement.

So we can imagine a world where the mobile network operators use both licensed and unlicensed spectrum in a coordinated fashion, using small cells and carrier aggregation, and the rest of the ecosystem continues to use and deploy traditional Wi-Fi at an ever-increasing rate. This world clearly separates LTE and Wi-Fi with very little coordination, so for subscribers to get the best possible user experience, a Wi-Fi service platform such as Devicescape’s will be critical. The Devicescape Service Platform can continue to manage the user experience on uncoordinated Wi-Fi while also managing the handoff to LTE where appropriate.

It will be interesting to see how carrier aggregation and uLTEA evolve over the next few years. This approach clearly indicates a new phase in tackling mobile broadband demands. I will discussing all of these topics on the June 3 panel of TIA’s Network of the Future panel: Carrier Aggregation in Licensed and Unlicensed Spectrum. It should be an interesting discussion.