I will follow William Webb, former director of Ofcom, UK telecommunications regulator where he heads the Spectrum Framework Review, the development of Spectrum Usage Rights, and British white space policy. As CEO of Weightless SIG, he leads the standard body for global M2M technology. He is now consulting. Webb has published 17 books and more than 100 papers on the spectrum and holds 18 patents. We discuss the challenges of spectrum policy including the latest Ligado controversy.
Radio spectrum policy continues to grow with market demand for new wireless services and innovation in spectrum efficiency. However, this progress sometimes clashes with the granting of historical rights to federal agencies who are not required to use their spectrum efficiently and are tied to technology that may be difficult and expensive to improve. These challenges are common in various countries. What can US policy makers learn from the UK?
Ofcom and telecommunications providers Arqiva see potential interference with terrestrial TV reception when 4G is introduced to the 800MHz band. They concluded that 2-3 million households will experience disruptions. Ofcom continues to advance, but with a large mitigation program in place to send filters to those affected. In practice the disturbance experienced is only about 1/100th than predicted. Indeed, in almost all cases where there is a terrible disturbance warning, but enlightened regulators push ahead also minimal disturbances are actually experienced. That reminds me a little of the predictions of the Y2K disaster 20 years ago – some problems were avoided from warnings, most never materialized, and some that happened were resolved quickly.
That reminds me of how the National Oceanic and Atmospheric Administration (NOAA) warned that the Federal Communications Commission (FCC) 24GHz continuing to kill the weather forecast or how the Transportation Department said that 5.9GHz and will cause a car accident. I do not regret the institution that cares, but institutional resistance to adopting new and better technology and technology can delay, if not hinder, the development of valuable new services.
The “change of use” decision is made by regulators with the expertise and authority to make it. The problem is they tend to resist risk. This is not a criticism like observation. After seven years at Ofcom, I understand why risk avoidance is inevitable. Ruling federal users tend to have far greater lobbying power than newcomers for political and sometimes financial reasons. The result is that many spectrum debates are biased towards the status quo. This manifests itself in time-consuming studies and expensive tests that accumulate one worst case assumption in another ending with a scenario that can cause interference, but it is very unlikely and can be reduced if it does.
Look what happened with Ligado. After a decade of research, revocation of spectrum access and bankruptcy, several expensive tests, and much heated debate, Ligado finally got permission from the FCC to start spreading 4G / 5G solutions in L-Band where GPS also resides. But now the US Department of Defense, one of the most obstinate incumbents, claims that many military systems will become inoperable. If this really happens, then this system is not suitable for certain purposes. Enemies, terrorists, or dissatisfied clearly can easily trap them with high-powered in-band jammers that are unfortunately very available, even on the Internet. Fragile systems must be replaced with solutions that use multiple satellite systems, have best-in-class filtering, have antennas that can resist interference, have complicated codes like cell phones, and more.
Resolving interference disputes should be easy with tests and studies, but it’s not. Why not?
Interference studies can guide potential problem areas, but there are many variables to consider. First, we need to decide what constitutes harmful interference, the regulatory standard observed throughout the world. Understandably, incumbents will claim that any change, however small, is problematic – they have no incentive to do otherwise. This tends to lead them to the requirement that noise floors can only be altered by a very small amount, if at all, if any, to the extent that the GPS community believes that noise floors outside their band must also be within their right to control. – effectively claims ownership of a band that is far wider than allocated, eliminating all norms in efficient spectrum management.
Disruption studies are only one piece of the puzzle. Services that might be affected – in this case GPS – will change. For example, the TV interference case discussed earlier assumes the existing TV receiver has poor filtering. But subsequent models, post cellular allocation, have been designed to accommodate 800MHz cellular interference and when cellular service is used for more than five years, the population of TV receivers tends to increase rapidly.
Incumbents are not interested observers and their subjective views must be treated with caution. Instead, we need to look at this study from the perspective of a country where the right balance is needed between allowing new services while not disturbing materially the present. To find out if disruptions are material, we need to see whether services – in this case usually location determination, although time is also important – are less accurate to reduce their efficacy in use. When doing research, we need to keep asking whether we make sense or whether we see the worst case that is very unlikely.
So technology generally improves faster than testing quality. What do policy makers need to know about GPS that will help in handling interference disputes?
Given the time span of the Ligado process, newer GPS devices may already have strong filtering with anticipated continuous improvements based on the decision. They also increasingly utilize satellite constellations, including the new European Galileo system, making them far more vulnerable to interference. Studies show that GPS and Galileo are sensitive to various types of interference, so if one is experiencing interference, it makes sense that the other is not. Many GPS devices, such as cellphones, fitness trackers, and many more are changed regularly, but even commercial, military, and precision GPS receivers have very limited life spans. Assuming that the current device does not change is usually an assumption that is too risk averse.
Despite the best efforts in overcoming risk aversion in initial decisions, it is almost guaranteed to be an excessive risk aversion. For example, to allay existing concerns, Ligado has agreed to a very low power level and a guard band that is almost unheard of (27MHz compared to typical values of 10MHz or less – for example a British escort band between cellular and TV at 700MHz is proposed 9MHz area as shown in Figure 2.1 below.).
The way to overcome this risk aversion is to allow the license conditions to be modified in accordance with actual evidence. When services are deployed, tangible impacts, as opposed to the impacts they model, can be studied. When it is clear that more interference can be tolerated, the safety tape can be reduced and the power level increased.
How much does it cost to avoid risk?
It is always easier to see the adverse effects on existing services than the negative impacts on new services in the future that do not yet exist. For example, if “white space TV” is quickly available, we might have an extensive Internet of Things network throughout the country, providing increased productivity and important quality of life. Imagine also if services like 4G, 5G and Wi-Fi6 were not permitted because of incumbent incumbents – our society, especially now with Covid-19, would be far worse. The Liberating Ligado Spectrum enables the innovative IoT and 5G networks to be used in spectrum bands with propagation far better than the 3GHz band suggested by others, allowing new services to be widely used more quickly and everywhere. The FCC made the right decision to balance risk avoidance with innovation.
I understand that the British military is actually paying fees for using their spectrum. Can you talk about the process for managing the settings and how they work? Can it work in the US?
The concept grew out of expectations that military users and other public sectors might have a greater spectrum than they really need. It was easy for them in the past to ask for a large assignment, and there was no incentive to give up – quite the opposite, no military planner wanted to be known as the person who returned the spectrum needed later. Determining the spectrum of what is of little value to them is almost impossible for outside observers because of the secret nature of its use. So instead, after commissioning an independent report, Ofcom decided to apply the annual fee to the British military as an incentive to return the spectrum thereby reducing costs. Once returned, the spectrum can be incorporated into a “market power” management model that allows flexibility in its use. Such an approach can work well in the US – the same underlying problems and drivers apply here as in the UK.