Today the wireless industry is gripped by debate over the deployment of LTE services in unlicensed spectrum. Qualcomm is the leading proponent and has recruited support from Telecom Equipment Manufacturers and Mobile Operators. Qualcomm has proposed three separate standards: LTE-U and License Assisted Access (LAA) are targeted geographically. A third proposal called MuLTEfire uses unlicensed spectrum for Enterprises, and does not require an LTE channel in licensed spectrum. Details of those proposals can be found elsewhere.

The goal of this article is to understand the motivations for LTE-U, and the reasons for the design decisions that have been made. In this article the technologies collectively will be referred to as LTE-U.

Almost everyone benefits by increasing the spectrum available for LTE. But “more LTE” alone does not explain why LTE-U is setup the way it is. Nor does it explain why some mobile operators support it and some don’t.

Here are some of the key points about LTE-U:

• LTE-U supporters include Qualcomm, Telecoms equipment manufacturers such as Ericsson, Samsung and Alcatel-Lucent, and Mobile Operators, mostly larger operators such as Verizon Wireless and T-Mobile.
• LTE-U and LAA require the network operator to offer LTE using licensed spectrum.
• Supporters claim LTE-U will co-exist nicely with WiFi and other users of unlicensed spectrum but are resisting calls for independent and transparent testing.

The behavior of the players in this drama can best be understood by considering their motivations. Let’s look at the players.

Qualcomm

Qualcomm collects revenue from manufacturers who license Qualcomm intellectual property (IP). Patents, in other words. These manufactures make both end-user devices such as smartphones and tables, and network devices such as Base Stations and eNodeBs. Every company selling LTE-capable devices pays royalties to Qualcomm. This strongly incents Qualcomm to increase the number of devices using LTE. Therefore Qualcomm’s primary motivation for LTE-U is to increase revenue. For Qualcomm “more LTE” means more revenue.

Qualcomm also collects revenue from WiFi chipsets. But the WiFi standard requires patent holders to license at reasonable rates. On the other hand, Qualcomm is free to license its LTE IP at whatever rate the market will bear. So net-net, Qualcomm makes more money from licensing its LTE IP than it does from licensing its WiFi IP. So “more LTE” nets Qualcomm more than “more WiFi”.

Telecom Equipment Manufacturers

Several of the large telecom equipment manufacturers (vendors) are supporting LTE-U, including Ericsson, Samsung and Alcatel-Lucent. To understand why, let’s consider their business models. The Telecom vendors make money selling hardware, software and services to mobile operators. (Being large companies they have other lines of business unrelated to LTE-U.) One significant source of revenue is the sale of capacity licenses. Capacity licenses are software keys that unlock additional capacity from equipment already in service. These licenses generally are cheaper than buying more equipment, but only a bit cheaper. The licenses are very low cost for the vendor, so capacity licensing is very high-margin revenue, and the vendors love it. The mobile operators, not so much.

The vendors also pay royalties to Qualcomm. Their support of LTE-U clearly indicates they stand to collect more capacity license revenues from their operator customers than they royalties they will pay to Qualcomm.

This capacity-licensing regime is less common and less lucrative for vendors in more consumer-oriented technologies such as WiFi. The Telecom vendors dearly want to increase these high-margin revenues and can do by expanding LTE into unlicensed spectrum. This is the reason LTE-U requires the mobile operator to also use licensed spectrum. It couples the technology to the Telecom vendors who sell kit for service in licensed spectrum, and more importantly, binds it to the lucrative capacity-licensing regime. But this is a simple design decision. There is no technological reason unlicensed spectrum must be paired with licensed spectrum.

Additionally, Qualcomm may have felt more industry support would be needed to make LTE-U an approved standard, and thus designed the standard for the benefit of those supporters. The Telecom vendors might have participated in this design decision. Requiring the use of licensed spectrum is primarily intended to get the large Telecom vendors on board and eliminate competition from smaller manufacturers. This is an important point.

Smaller manufacturers have been successful selling equipment in unlicensed spectrum using WiFi. Over the last several years much of the growth of mobile data usage has been served not by LTE networks, but by WiFi networks. This is called “mobile data offload”. Mobile Operators like this because WiFi networks are cheaper to build and operate than LTE networks. Mobile data offload also has the effect of shifting revenue from the Telecom vendors to the WiFi vendors. Increasing LTE service using LTE-U will reduce or slow this revenue shift by keeping more mobile data on LTE, rather than WiFi.

So to summarize the reasons Telecom vendors support LTE-U:

• To increase capacity-licensing revenues
• To reduce competition from WiFi equipment vendors

Mobile Operators

If LTE-U generates more vendor revenue paid by mobile operators, why would some mobile operators support LTE-U? Why not deploy more WiFi, which is cheaper than LTE, to increase mobile data offload?

We said earlier that WiFi network were cheaper to build and operate than LTE networks. Part of the build saving comes from avoiding spectrum licensing fees, and part from the less odious vendor capacity-licensing regimes. WiFi networks are becoming increasingly capable, too.

• These days WiFi is built into almost every electronic device imaginable.
• Many Internet of Things (IOT) devices connect to the Internet via WiFi.
• The Hotspot 2.0 standard is simplifying WiFi interoperability with mobile networks.
• Voice over WiFi, VoWiFi, is becoming more solid.
• The WiFi standard, 802.11, continues to evolve and mature.

The lower cost of WiFi networks and increasing performance has led to the rise of WiFi-only mobile networks. In the U.S. the operators Freedom Wireless and Republic Wireless are already making inroads with WiFi-first service. This means price competition for incumbent mobile operators. By extending LTE into unlicensed spectrum, mobile operators can avoid the high fees for licensed spectrum.

Worse, the unwillingness of the LTE-U group to submit to independent and open coexistence testing raises concerns that LTE-U might degrade WiFi service.  Degradation of WiFi could be a hidden agenda of LTE-U, to eliminate competition and gain more spectrum for LTE. The notion may sound a bit like a conspiracy theory. But that doesn’t mean it can’t happen.

Regarding the increased capacity licensing costs. Large mobile operators can demand better terms from their vendors than small mobile operators. They can do this through their greater scale, and by playing vendors against each other, preferring the vendor giving the best terms. Smaller operators are less likely to have multi-vendor networks, so playing them against each other cannot happen. Of course, there is the “Nuclear Option”, the threat of a complete network swap to another vendor, though this is less common. Either way, large operators can get better prices than smaller operators.

So some large operators might believe that benefits of LTE-U, including crowding out WiFi-only service providers, offsets the potential for increased licensing costs paid to the Telecom vendors.

Summary

So lets recap the motivations of companies supporting LTE-U. Adoption of LTE-U:

• increases Qualcomm’s patent royalties
• increases Telecom vendors’ capacity licensing revenues and thwarts WiFi-centric equipment competitors
• Reduces downward pressure on Mobile operators price from low-cost and WiFi-first competitors

Note that this is not to say that LTE-U is a bad thing. It’s not. But as presently designed, it has characteristics that are unacceptable. It tilts the playing field too much in favor of the proponents. Let’s look at how LTE-U could be changed to make it better for mobile subscribers.

Remedies

Qualcomm may have always assumed that open and independent coexistence testing with other users of unlicensed spectrum ultimately would be required of LTE-U.   By initially resisting calls for such testing Qualcomm might hope to later claim they’ve made a concession when testing finally is required.   Regardless, LTE-U should be subjected to rigorous testing before being allowed in the wild. This lessen should have been learned in the U.S. from the recent experience of LightSquared using satellite spectrum that was polluted by poorly designed GPS equipment, rendering it completely unusable for its intended purpose.

1. LTE-U must be rigorously tested by an independent body in an open and transparent manner, to assure peaceful coexistence with other users of unlicensed spectrum.

Qualcomm might hope that agreeing to independent and open testing will be seen as a concession that allows other aspects of LTE-U to be retained, such as their patent royalties. But use of unlicensed spectrum is a privilege. Qualcomm cannot be allowed to extract the same royalties from unlicensed spectrum as their licensed-spectrum technologies.

2. All intellectual property to be used in unlicensed spectrum must be licensed under Reasonable And Non-Discriminatory terms, just like WiFi itself. In other words, Qualcomm will not be allowed to collect the same patent royalties from devices using unlicensed spectrum that they collect from devices using licensed spectrum.

Requiring mobile operators to use licensed spectrum alongside LTE-U is anti-competitive. So let’s eliminate that. This will allow anyone to offer LTE-U service.

3. Possession of licensed spectrum will not be a requirement for using LTE-U.

Lastly, the growing capabilities of WiFi could foster development of WiFi-only operators. If so, a mix of small and large operators can be expected. Inter-operator and inter-technology roaming should be encouraged.

4. WiFi network operators must allow “roaming” or inter-operability with other operators on RAND terms. This approach has been implemented in the US for voice roaming among mobile operators. It helps maintain a healthy competitive landscape by lowering costs for operators who lack the size to derive economies of scale.

The guiding principal for using unlicensed spectrum is that barriers to entry should be as low as possible for compliant technologies.

In the US the FCC, and relevant regulatory bodies in other regions, should demand the remedies described above. And ultimately, the LTE-U standard should be stewarded by an impartial body having members from all aspects of the mobile business.