Will LTE in Unlicensed Spectrum Unlock a Vast Store of Mobile Broadband Capacity?


By Peter Rysavy, Rysavy Research, for MIMOWorld.com:

The barrier separating Wi-Fi in unlicensed spectrum and cellular in licensed spectrum has been breached—with potentially far-reaching consequences for the mobile broadband industry. In late 2013, Qualcomm and Ericsson presented a proposal to the Third Generation Partnership Project (3GPP) for extending LTE to unlicensed spectrum in the 5 GHz band. The goal is to create a supplementary 20 MHz channel that can be combined with channels in licensed spectrum using LTE Advanced’s carrier aggregation mechanism.

The advantages to cellular operators are significant. Base stations, particularly in small cell environments, would have considerably more bandwidth available, delivering higher average throughputs to LTE users. Plus, operators could avoid the complexities of handoffs between Wi-Fi and cellular.

Another advantage is that LTE networks use spectrum more efficiently than Wi-Fi networks, especially under heavy load. That’s because LTE, unlike Wi-Fi, is a managed network. For example, LTE employs sophisticated scheduling algorithms for sending data to mobile users. By favoring the devices that have the best signals at the moment, the physical layer can use more aggressive modulation and less aggressive error correction modes, resulting in higher average spectral efficiency.

In fact, the benefits for cellular operators seem almost too good to be true.

Still, differences have emerged at recent 3GPP meetings showing that LTE in unlicensed spectrum is not as simple and uncontroversial as it might at first appear. For example, if licensed networks are allowed to borrow capacity from unlicensed spectrum, then there will be less capacity left for unlicensed users. Interestingly, handoffs from cellular to Wi-Fi would have the same effect but for some reason don’t evoke as many objections.

There is a more serious problem, though. Wi-Fi technologies have always been designed from the ground up to co-exist with each other. For example, if two Wi-Fi networks within hearing distance of each other wish to use the same radio channel they may do so using media access protocols designed to enable sharing. Each station must wait until the channel is clear before attempting to transmit. And even then they may only do so after waiting a random period. This prevents multiple stations from pouncing on the channel at the same time.

In contrast, unlicensed spectrum rules in some countries permit LTE operators to preempt Wi-Fi users on any channels the licensed operators choose to occupy. This is one source of disagreement among 3GPP participants, because some operators have invested heavily in public Wi-Fi infrastructure. Such a heavy-handed approach would likely cause a public outcry, so engineers are looking at changes that could be made to LTE in order for it to better coexist with Wi-Fi in the 5 GHz band.

Assuming that a good neighbor solution is found that is acceptable to both the cellular and Wi-Fi industries, operators still face a fundamental network architecture question. Namely, what is the best way to deploy millions of small cells and unleash the next wave of mobile broadband capacity?

Let’s consider a hotel environment. Is it better to deploy multiple LTE small cells throughout the property or multiple Wi-Fi access points? Today, Wi-Fi has the advantage because Wi-Fi access points are less expensive and Wi-Fi can serve as a “neutral host” solution.

A neutral host is one that enables multiple operators to share the same infrastructure equipment. Using a Service Set Identifier (SSID), a Wi-Fi access point can advertise itself and offer service on behalf of multiple networks. Assuming appropriate network integration, anyone can deploy a Wi-Fi network that appears to be part of a cellular network, a hotel network, or any other network that has a contractual relationship with the infrastructure owner. On networks adhering to the Hotspot 2.0 standard, existing subscribers are automatically connected and given access to their usual services and features.

In contrast, cellular infrastructure equipment is traditionally designed to serve a single operator. Therefore, if multiple cellular operators wanted to provide service throughout a hotel using small cells, then each cellular operator (including any using LTE in unlicensed spectrum) would have to install its own equipment, an approach that is at best inefficient and potentially quite problematic.

There is an alternative worth exploring: the creation of neutral-host LTE small cells that serve multiple operators. Vendors are developing this capability, but whether cellular operators will be willing to share network access equipment with their competitors remains to be seen. An obvious extension of this idea would be neutral-host small cells supporting both Wi-Fi and LTE operation in unlicensed spectrum. The ability to lease both LTE and Wi-Fi capacity might give third parties the incentive they need to aggressively deploy small cells.

The success of LTE operation in unlicensed spectrum will depend on how aggressively small cells are deployed and the growth of public Wi-Fi infrastructure—two aspects that are evolving rapidly and somewhat unpredictably.

Developers of LTE in unlicensed spectrum must also take into account how Wi-Fi operation varies from country to country. For example, in the U.S. operation in the 5 GHz band requires APs that support Dynamic Frequency Selection (DFS), the ability to move to a different channel when the presence of government radar is detected.

LTE operation in unlicensed bands is going to take time—if only for the reasons already discussed. The technology is being considered for 3GPP Release 13, a version of the standard that is not likely to be completed until 2016. Consequently, the earliest equipment would become available is 2017, though 2018 is more likely.

At this early stage, it’s hard to predict how many operators will adopt the technology, but the benefits are too compelling to ignore. The big question is: Will LTE in unlicensed spectrum be just another niche solution or will it play a dominant role?

Peter Rysavy is the president of Rysavy Research, http://www.rysavy.com, a company that has specialized in wireless technology for the past 20 years.