There’s been quite a lot of activity on LTE-U and LAA in recent months, with a major joint standards meeting in Beijing that included both Wi-Fi and cellular communities. Some chipset, RAN vendors and large operators have been backing the use of LTE within the 5GHz unlicenced band (mostly used today by Wi-Fi), specifically as an overflow or supplementary downlink capability - so only used for larger downloads/video etc. when the normal cellular bands are congested. The Wi-Fi industry has been very concerned but remains open to reaching a consensus.
Recent progress was highlighted by a well publicised 3GPP workshop in Beijing at the end of August, with close to 300 attendees including from the Wi-Fi community. It’s the first time these two communities have really sat down at the same table to work out their differences. The IEEE (who set the standards for 802.11 Wi-Fi) and 3GPP (which defines cellular standards) both said they’d like to co-operate. The Wi-Fi Alliance and Wi-Fi Broadband Alliance (WBA) also attended. The focus of the meeting was the use of LTE in the 5GHz unlicensed band currently used by Wi-Fi (amongst others).
IEEE 802.11 committee is open to sharing spectrum fairly
There are some in the Wi-Fi industry who would very much prefer that the use of LTE in unlicensed spectrum wasn’t being considered at all. Or at least, that it was constrained to a non-Wi-Fi band such as 3.5GHz. But their standards body, IEEE 802, is open to the possibility provided that those using LTE adopt an “802.11 like” access mechanism - often termed Listen Before Talk. They are ready and willing to work with 3GPP in a truly collaborative manner.
Wi-FI Community stand firm on Listen Before Talk feature
Wi-Fi has succeeded largely because of a polite sharing mechanism called Listen Before Talk with truncated exponential back off - where congestion is detected, transmitters wait for longer and longer periods each time before resending. It’s worked very well for 15 years and with around 5 billion devices in use (similar numbers to cellular) looks likely to stay. While innovative new approaches must always be considered, the timescales planned for LAA make is very unlikely that consensus for anything new would be achieved. There simply isn’t the time to conduct extensive technical studies and gather adequate proof.
3GPP TR 36.889 is recommending the use of a Category 4 Listen Before Talk mechanism, which has many similarities to 802.11. The term “802.11 like access” is being used to mean Listen Before Talk. Category 4 requires different licensed frequencies for the macro and small cell layers, and it’s noted that few operators assign dedicated small cell bands today. Carrier Aggregation has meant less available spectrum to set aside, not more.
Focus on downlink first
This technical report also recommends a focus on downlink only. That’s not a problem since LAA is intended to provide additional download capacity as a supplement to regular cellular service on licensed bands only when required. LTE already has the advantage of a more robust uplink, and the fact that most traffic is downlink today (although the proportion is changing), this brings the quickest benefits. For voice calls, I've heard from several sources that it's difficult to handle more than eight concurrent voice sessions per Wi-Fi access point because they can't be prioritised and reliably avoid jitter and dropout. Inherently LTE has that scheduling built-in, allowing it to handle hundreds of concurrent calls per sector.
Industry Consensus beats Regulatory Imposition
The IEEE would much rather that those in the industry come to a consensus rather than have an arbitrary solution imposed by regulators. Indeed, regulators often prefer that stakeholders collaborate with a recommendation that they can then adopt.
Although 3GPP does have many liaisons and partnerships with other organisations, it doesn’t have all the mechanisms in place to facilitate a collaborative approach. Some procedures will have to be put in place to work with the IEEE on this.
3GPP has a work program which schedules LAA in their Release 13, due out March 2016. That’s quite a tight timescale. In addition to 3GPP committees, the Wi-Fi Alliance will be hosting some workshops during November.
Meanwhile, some industry players are proceeding apace with LTE-U which deploys LTE in the 5GHz band without any modifications. This is legal in some countries, such as the US/Korea/Japan, but not in many others such as Europe. This would pretty much kill any Wi-Fi traffic on whatever frequencies it used because Wi-Fi backs off nicely while LTE (without LBT) won’t. Wi-Fi would then have to find other bands to use instead and dynamically switch.
Ericsson announced their latest major RAN software release 16A, which includes LTE-U capability as a step towards LAA. They’ve said earlier this year that their small cells would support LTE-U before end 2015, effectively just operating LTE in the 5GHz band.
But would complementary use of Wi-Fi be quicker to provide greatest community benefit?
Peter White of Rethink Research published a long and insightful opinion piece discussing the tradeoffs and still believes LAA has some fatal flaws. He points out that it will take quite some time for devices and small cells to be able to make use of LAA. There will be many operators who already make heavy use of Wi-Fi and would not want to see it deprived of full capacity or efficiency. The service may become restricted to a limited set of operators and smartphone users.
Joe Madden of Mobile Experts points out that LAA would really only benefit those operators already making large investments in Small Cells.
Others point out that making full use of Wi-Fi requires additional core network gateways, security mechanisms, management tools and other infrastructure. It’s just much easier for operators to use the same LTE network at a different frequency. The primary use case is for public area use, mostly indoors. Huawei rules out its use for residential small cells.
The LAA initiative still has quite some journey to make, but it’s definitely making progress and one to watch over the coming year ahead.
TR 36.889 - 3GPP's initial technical report on the feasibility of LAA