The Trade-Off between Innovation and Regulation in Unlicensed Spectrum

| Andrew Myles

Way back in 1985, the FCC in the United States made the momentous decision to open up the 2.4GHz band for unlicensed use. The 2.4 GHz band is often called the “junk band” because it also contains emissions from everything from RF welders to microwave ovens. The associated rules imposed few restrictions and opened up the band as one great big experimental platform that would determine if unrelated users could coexist together in unlicensed spectrum in a useful way.

The freedom to communicate wirelessly without a license attracted a wide variety of potential users. In 1990, the IEEE 802.11 Working Group formed to develop a new WLAN standard to take advantage of this amazing opportunity of free spectrum for everyone. It took seven years to complete the first ratified version of the IEEE 802.11 standard, but the result provided the basis for the socioeconomic goliath now known as Wi-Fi. In 2017, more than half the world’s IP traffic is carried over Wi-Fi, making use of more than nine billion Wi-Fi devices, with more than three billion new devices sold annually. Wi-Fi initially was focused on connecting laptops to the Internet. It is now used for everything from industrial machines to laptops to phones to bathroom scales. As a guest worker in Dubai recently noted to the author, “Wi-Fi is life!” because it allowed her daily access to her family in the Philippines.

One question often arises: why was the FCC’s experiment to allow anyone to use the 2.4 GHz band so successful? Generally, if you allow anyone to do anything, they usually do, and chaos often ensues. If chaos had occurred in the 2.4 GHz band, Wi-Fi today would not allow “anyone, anytime, anyplace to set up a network that just works well enough,” adding hundreds of billions of dollars to the worldwide GDP annually. On the other hand, if the experiment had led to chaos, then the global community would have not been any worse off than the status quo; it would have simply been a missed opportunity.

One possible answer is that it was just chance that the unlicensed wireless market generally converged without much oversight on a single solution based on the IEEE 802.11 standard, which uses a relatively conservative Listen Before Talk (LBT) protocol, with distributed control. This approach inherently gives everyone roughly equal access. It was also probably fortuitous that the Wi-Fi Alliance formed in 1999 to prove interoperability, using[KZ1] certification tests that effectively used sharing as a proxy for interoperability. The efforts of the IEEE 802.11 Working Group and the Wi-Fi Alliance working with a homogeneous sharing mechanism meant that fair coexistence in the 2.4 GHz band came for free.

Wi-Fi is not the only wireless network using the 2.4GHz band. Both Bluetooth and Zigbee have a significant presence in the band too. However, again the global community was fortunate. Rather than attempting to make quite different systems coexist, both Bluetooth and Zigbee were designed so that they could, for the most part, avoid operating in the same channels as Wi-Fi. Instead, they often focus on the space between or beyond Wi-Fi channels in the 2.4 GHz band.

Wi-Fi’s early success meant that the spare capacity of the 2.4 GHz band quickly diminished. Regulators around the world had the foresight to recognise this possibility and started identifying additional spectrum in the 5 GHz band for non-licensed use from about 1998. The available 5 GHz channels have varied by country and over time, but one common characteristic has been that Wi-Fi often has to share the channels with primary users such as satellites and radar. The need to share with primary users means that Wi-Fi is required to limit its output power in some channels to avoid interfering with satellites, and to not use some channels at all when radar is detected. Deferring to primary users is in everyone’s best interest; no one wants Wi-Fi to interfere with the weather radar at the airport that is helping your plane to land safely. The fortune of the 2.4 GHz band has extended into the 5GHz band insofar as fair coexistence between users of Wi-Fi systems can almost always rely on the distributed sharing enabled by the LBT protocols used by the IEEE 802.11a/n/ac amendments.

A few years ago, operators using LTE in licensed spectrum discovered the benefits of unlicensed spectrum: access to capacity they did not have in licensed spectrum, and for free! Some operators utilized Wi-Fi with various degrees of integration with their licensed systems. More recently, efforts have started to focus on the use of LTE-based mechanisms in the 5 GHz band. The LTE-U Forum developed LTE-U, 3GPP developed LAA, and the MulteFire Alliance developed MulteFire. These new systems have raised serious coexistence questions for Wi-Fi and more broadly for the global community that depends on Wi-Fi’s socioeconomic benefits.

LTE-U is a system that decides on what is “fair” on behalf of the Wi-Fi systems operating in the same channel. This centralized approach is quite different from the distributed approach used successfully by Wi-Fi for so many years. The threat to fair coexistence led to significant debate in the United States, with the FCC and even committees of Congress becoming involved. It was a very painful process for all parties, particularly the FCC, which preferred that the market decide as it had in the past. Eventually, the Wi-Fi Alliance facilitated the development of a test plan that would identify the worst violations of fair coexistence by LTE-U. While this is helpful, it provides no guarantees because its use is not mandatory. Ultimately, LTE-U is not expected to have a significant market impact, so any associated coexistence issues will be limited.

LAA (and MulteFire) is a system that is based on an LBT protocol, like Wi-Fi. Interestingly, this design choice was heavily influenced by European regulatory requirements for LBT and a desire to deploy LAA globally. These same requirements mean LTE-U cannot operate in Europe, which is part of the reason LTE-U is unlikely to have longevity in the market. The use of LBT by LAA should result in some degree of fair coexistence with Wi-Fi. Unfortunately, it is not guaranteed because LAA uses a different form of listening to Wi-Fi. Whereas Wi-Fi detects preambles at -82 dBm and energy at ‑62 dBm to determine whether the medium is busy, LAA uses only energy at -72 dBm. Because this asymmetry will inevitably cause problems in certain environments, it is the subject of continuing discussions by standards organizations and regulators.

Going forward, there are a number of unanswered questions. What is the correct balance between allowing innovation of the nature that allowed Wi-Fi to emerge in 1999 and protecting the socioeconomic benefits of Wi-Fi that exist today? Is the laissez-faire approach of the United States or the light touch regulation of the Europeans better? Will the use of different sharing mechanisms allow for sufficiently fair coexistence between LAA/MulteFire and Wi-Fi? If not, is there a case for even tighter regulation? Or should we just let innovation rule, always letting the market decide?

In 1985, the FCC started a very successful experiment to make use of the 2.4 GHz band. In 2017, it appears a new experiment is being run in the United States and Europe to find the right balance between innovation and regulation in unlicensed spectrum. However, in contrast to the original experiment that was conducted in an environment with no significant stakeholders, the latest experiment brings together two economic juggernauts: the Wi-Fi industry and the LTE industry. The IEEE 802.11 Working Group, 3GPP, Wi-Fi Alliance, MulteFire Alliance, LTE-U Forum, and regulators are all stakeholders, along with everyone who uses Wi-Fi today. Let’s just hope this experiment turns out as well as the original, with only winners. Only time will tell!

Dr. Andrew Myles has a B.Sc./B.E. (Hons. I) from the University of Sydney and a Ph.D. from Macquarie University. He has worked in higher education, industry research, management consulting, and industry for 30 years. He is currently employed at Cisco. For the last 17 years, his focus has been on standards work related to Wi-Fi, including in the IEEE 802.11 Working Group, Wi-Fi Alliance, ISO/IEC JTC1/SC6, and ETSI BRAN, both in technical and governance/leadership roles. When asked what sort of engineering he does, he now answers that he is a political engineer.