As operators across the Asia-Pacific region continue to roll out LTE networks, as well as enhance existing deployments through carrier aggregation, further technology upgrades using unlicensed spectrum are becoming available. To increase capacity and peak data rates, operators have the option to either combine LTE and Wi-Fi or use LTE in unlicensed spectrum.
These technology options are predominantly aimed at localized traffic hotspots. Because the unlicensed frequencies being used will initially be in the 5-GHz band, these solutions are particularly applicable for small cell deployments. However, aggregating LTE and Wi-Fi could involve using existing Wi-Fi access points combined with an LTE macro network.
Combining LTE and Wi-Fi could involve using both technologies on the downstream and upstream, or improving capacity by using Wi-Fi on the downstream and cellular on the upstream. The aggregation of LTE and Wi-Fi could occur at various levels of the network. Some of the options available include PDCP (packet data convergence protocol) modem level aggregation, IP layer aggregation and MP-TCP (multi-path TCP) aggregation.
See Also
LTE Insights October 2015
A further set of technologies involve using LTE itself in unlicensed spectrum. Operators could deploy LTE-U, the pre-standard version of LTE for unlicensed spectrum developed by Qualcomm, or use LAA (Licensed Assisted Access), which is presently being standardized by the 3GPP.
Coexistence challenge
The development of LTE in unlicensed bands has focused on aggregation with licensed LTE spectrum, which acts as the base that enables the unlicensed spectrum to provide an additional capacity boost. LTE in unlicensed spectrum can add capacity to licensed spectrum by using a supplemental downlink for the downstream, or in later versions, carrier aggregation for the downstream and upstream. When using LTE in unlicensed spectrum, end-user devices and small cells will require new chipsets supporting LTE in the 5-GHz band.
Using LTE in the unlicensed 5-GHz band will require it to coexist with Wi-Fi, which raises a number of issues. In some markets, existing 5-GHz Wi-Fi deployments must be protected through the use of “listen-before-talk” mechanisms. In these cases, operators must wait for the standardization process of LAA to be completed, which is expected in March 2016. However, in other markets there is no regulatory obligation for listen-before-talk mechanisms, and existing Wi-Fi deployments can be protected using mechanisms based on CSAT (carrier sensing adaptive transmission) available in Qualcomm’s LTE-U technology.
In Asia specifically, listen-before-talk mechanisms are not required in some markets - notably South Korea, China, and India - which makes these markets prime targets for deployments of LTE-U. In other markets - for example, in Japan as well as in Europe - listen-before-talk mechanisms must be used to protect existing Wi-Fi deployments, making LAA the requisite technology.
It’s possible we will see the commercial launch of LTE-U systems in the second half of 2016, while operator launches of LAA may begin from mid-2017 onwards. This means that markets like South Korea will be among the first to deploy LTE in unlicensed spectrum - indeed, Korean operators are already trialling LTE-U. In May 2015 LG UPlus carried out a demonstration of LTE-U achieving speeds of 600 Mbps. This demonstration used 60 MHz in the unlicensed 5.8-GHz band and 20 MHz of licensed LTE spectrum.
Operators can also choose between various ways to aggregate LTE and Wi-Fi in order to boost capacity and peak speeds. IP and MP-TCP layer aggregation do not require new hardware, and in order for subscribers to be able to use these technologies, operating system upgrades can be made available to them.
Modem-level PDCP aggregation requires a new chipset in end-user devices. Whichever method of combining LTE and Wi-Fi is chosen, these technologies would have a quicker time-to-market compared to using LTE in unlicensed spectrum. However, major players such as Apple may hold the opinion that these technologies cede too much control to operators, and may be unwilling to support OS upgrades or procure chipsets supporting modem level aggregation.