Mind Matters Natural and Artificial Intelligence News and Analysis

# Radio Is Changing From Hardware to Software: New Policies Needed!

Perhaps surprisingly, theology may play a role in determining how radio will be used

As I noted in an article earlier this week, dynamic spectrum access networks (DySPAN), implementing policy-based radio, are defining the future of wireless. The new networks are based on ontologies.

Ontology is the study of the essential nature of a subject. It identifies the objects or variables that are involved. It then identifies the relationships between them. An ontology identifies the variables and how they relate to one another. That knowledge is then used to develop solutions.

The initial ontology focuses on the variables that may result in one radio service interfering with another. The subject is the transmission level. The purpose is to improve coexistence. The goal is to get transmitters to share the same spectrum with a minimum of interference. To improve coexistence, it is essential to have an accurate understanding of the physical layer. All the variables must be understood. Once understood, they must be managed so that reliability is maximized and the potential for interference is minimized. In this area, the topic being explored is what factors influence the level of interference between transmitters and how do those factors relate to one another?

However, the power of ontological thinking is not limited to the physical layer. It can be applied in multiple ways. It has an even greater impact when applied to higher, more abstract levels. Every time a new variable is included, the ontology becomes more complex. With every expansion, new variables become relevant. The relationship of those new variables to others must be understood and then managed. An easy example is to include the source or destination of a transmission. What if all messages to 911 get priority routing? Just adding that one variable makes the web of relationships significantly more complex. But including it also creates the potential for new benefits.

The power of ontology use on multiple levels is becoming more apparent. The effort may have originated in a need to improve the efficient use of spectrum but the potential for other benefits soon becomes clear. Ontological thinking can also be used to determine what applications AI and policy-based radio are developed to serve. A product developer might ask why some priorities for new development are followed. The answer will often be that these priorities are what is most important to the developer’s best customers. The flaw in that logic is that the new design is being optimized for the best customers in the past. Shouldn’t the developer be optimizing for the best customers in the future? The needs and priorities of customers in the past may have been vastly different. The needs of past customers may even conflict with the needs and priorities of future customers. Applied at this level, ontological thinking helps create a business plan that takes a company to its best future. Understanding the participants and relationships between them helps a company be future-oriented.

#### A Theology of Spectrum Policy

At the highest level, a theology of spectrum policy is needed. The use of the term “theology” may seem odd when talking about spectrum management. However, there is a repeated pattern in which the final decision is driven by values. That raises the question of where those values come from. Which values are to be given priority over others? When there are, as is often the case, conflicting legitimate concerns, how are those conflicts resolved? The answer to that depends on ones understanding of the nature of the universe and the character of God. In this context, God is understood as “the ultimate decision maker.”

There can be theistic and non-theistic approaches to such a question but either way, they are a theology or, alternatively, a philosophy. Whether the origin is theistic or non-theistic, values arise from one’s concept of the nature of the universe and the characteristics of the ultimate decision-maker in that universe. Finding optimal consensus solutions often means implementing compromise solutions that recognize the legitimacy of conflicting values and assigns them relative weight in the chosen solution. Developing working solutions that have broad support is facilitated if the values of those participating are clear and understood by all participants. It is not necessary that everyone agree or share the same views. But it is helpful if the participants understand each other’s perspective.

Governments use their values to determine the public interest. Spectrum regulators are delegated the responsibility for protecting the public interest. The job of the regulator is to understand the risks and regulate the use of spectrum in such a way that the public interest is adequately protected. Note that the public interest is to be adequately protected, not perfectly protected. There will always be risk. At times, the residual risk will be realized. For spectrum, there will always be some interference. The question—and it is a value judgment—is what level of protection is adequate? It is not uncommon to find that the greatest conflict centers on the acceptable level of risk.

Conflict often occurs when cost and benefits are separated. If I am being asked to accept the cost so that you can get a benefit, it is only natural that we will disagree. With spectrum management, the separation of cost and benefits is relatively common. Understanding the underlying relationship allows creative solutions to be applied. Suppose you want to start a new wireless service but your new service will cause interference for some of my customers. Let us assume you will make $100M/year with your service but I will lose$1M/year because of the interference. Once we understand this, someone might suggest that you pay be $10M/year. Now I am quite happy to lose$1M because of your interference because you are paying me $10M/year for that loss. You will be happy to pay me$10M/year because it lets you make \$100M/year. The key is understanding the ontology that is at work. In this example, there is an economic solution to an interference problem.

Increasingly, the source and content of a transmission will become part developing a transmission plan. Most of us would give an emergency call to 9-11 priority over transmissions for entertainment. Here is another level at which values are the ultimate deciding factor. It is unlikely that any of us will benefit from priority being given to a specific 9-11 call. Most of us would recognize that we or someone we care about may someday be the one to benefit. We would support 9-11 calls being given priority routing.

That is not a technical decision. It is a decision based on our values.
It is not just the source of a message that should decide its transmission priority. Of a hundred messages sent by a senior executive or commanding officer, there are probably only a few that deserve priority routing. In addition, the message content may impact the priority of other messages. A commander ordering a rescue mission should certainly be given priority but so should the spare parts request for one of the aircraft assigned to that mission. In an ontologically based network, that can reason about the message content, the network itself could adjust the relative priorities of related messages.

#### Funder’s Interest

We can apply this kind of system analysis to the direction of wireless development. There is a natural priority to serve the interest of those funding the work. The tongue-in-cheek Golden Rule, “He who has the gold rules”, is as true for technology development as for any other field. If the funder of development is a business, the natural priority will be to develop technology that maximizes the profits of that business. If the funder is the military, military priorities will be served. While such priorities are understandable, they are not necessarily the right values to apply from a whole of society viewpoint. Regulators were established long ago to balance national spectrum policy with narrower priorities, such as maximizing the profits of a specific business.

It is the job of the Congress to consider the interests of the whole of society. It then delegates implementation to the FCC. If the FCC responded only to current business interests it would have never allowed unlicensed spectrum. Services like WiFi and Bluetooth would never have been developed. The reason they would not exist is that they would not have had spectrum in which to operate. No specific business or even industry segment was served by allocating some spectrum to be unlicensed. But that decision has stimulated tremendous innovation and the formation of entirely new industries.

The decision to allocate some spectrums for unlicensed use and other spectrums for licensed use is ultimately based on a consensus theology or worldview. It arises from a belief in unknowable innovation that will arise, given an opportunity. It arises from a belief that the rules should allow a David to sometimes beat a Goliath. These decisions are based on values but are informed by technology. The technological possibilities guided the decisions. The ultimate decisions are based on consensus values, as codified by legislation, and implementing regulation.

#### Multi-Disciplined Development

Work in wireless communications increasingly requires a multi-disciplined skill set. This is even more true as policy-based radio advances. Development engineers must command both digital and RF design. They must be skilled in software programming and hardware development. They must understand fields and propagation, antenna design, risk assessment, risk management and a variety of other topics. As discussed earlier in this article, it is important that they have training in theology and ethics so that they can clarify the values they are implementing with their engineering. They will need skill in identifying conflicting legitimate concerns and arbitrate optimal solutions.

A wide array of skills is needed by those who help create the future of wireless. A specific device must be designed, tested, and put on the market. All parts of the product must operate as intended, including the digital circuitry and the RF components. The software in the product must operate reliably. However, even before there is a project there must be a business plan that justifies investing in the project. Before that there must be a path through the regulations that allows the product to operate legally.
As higher-level ontologies are applied, additional variables are incorporated into the decision-making. These require skills in these additional areas.

#### Conclusions

AI and wireless communication have a complex, interesting, and exciting future ahead of them. The initial stages of implementing policy-based radio used limited and more deterministic decision-making. The sophistication available today is vastly different from past hardware-based radios. As additional variables are brought into the decision-making, the complexity grows exponentially. With that complexity comes the necessity of understanding the values being served. By what measures is the next generation system superior to the current generation? Clarifying those values requires skill in both understanding the motivating worldview and crafting optimal compromises with other legitimate but conflicting values.

To work effectively in this emerging world requires a wide arrange of skills. In the past many of these skills were separated into different fields of study. Hardware design, software programming, RF and propagation, theology and philosophy have been traditionally considered to be different study subjects. Those designing future wireless communications will need skill in all these areas and understand how they interact in the DySPAN systems of the future.

You may also enjoy: We are running out of radio spectrum but AI might help.