The Federal Communications Commission (FCC) has issued a Notice of Inquiry (NOI) titled “Advancing Understanding of Non-Federal Spectrum Usage.” The item seeks information about the use of wireless spectrum by entities regulated by the FCC, which includes commercial users and non-federal public safety agencies.

The agency’s goal is to understand actual spectrum use and find better ways to manage spectrum, including improved mechanisms for spectrum sharing. Over the past few decades, the FCC has become more flexible in allowing several types of users to share frequency bands, including with the use of novel spectrum sharing database systems. These decisions to open frequency bands to more shared spectrum use often require the agency to determine the legitimacy of claims about spectrum scarcity or the ability to successfully share spectrum with different user groups. Thus, the FCC seeks to understand “how we could take advantage of modern capabilities for doing so in a cost-effective, accurate, scalable, and actionable manner.” And in particular, the FCC notes that the “the burgeoning growth of machine learning (ML) and artificial intelligence (AI) offer revolutionary insights into large and complex datasets” which then could be used to better understand the nature of current spectrum use.

Chairwoman Rosenworcel explained, “[w]ith demands on our airwaves growing with the internet of things, we want to better understand spectrum utilization in geography, frequency, and time. This is the kind of data that could help make our policies smarter and more effective. It could also help support new cognitive abilities that could teach our wireless devices to manage transmissions and avoid harmful interference on their own. In other words, smarter radios using AI can work with each other without a central authority dictating the best of use of spectrum in every environment.”

As the FCC details in the NOI, the federal government (including scientific research agencies such as NIST and NSF) has engaged in various efforts to collect data on spectrum use. Key questions raised in the NOI are as follows:

• Defining Spectrum Use: How can the FCC not just measure but also define and express overall spectrum capacity of a given frequency band? The NOI suggests that components of this metric could include the number of devices in use or the power levels or receiver capabilities of the relevant wireless systems.
• Band Specific Considerations: Should spectrum use be measured differently depending on what services access a frequency band? And should that be reflected in the geographic area over which use is measured?
• Data Considerations: The bulk of the NOI focuses on data related to spectrum use, from what is available to the agency currently to what should or could be collected. In terms of obtaining data, the FCC seeks information on the cost and burden of specific data collections; standardization; and technical accuracy. It also asks about approaches to data collection, approaches, such as: (1) crowdsourcing, (2) external data sources, (3) modeling, and (4) direct observation.
• Other Considerations: The NOI additionally puts forth a set of questions regarding data privacy and the security of the data that the FCC may collect or use, asking about implications from the Electronic Communications Privacy Act as well as the Communications Act.

Finally, the FCC seeks comment on what activity should follow the NOI. One suggestion is the issuance of a Policy Statement setting out best practices for evaluating spectrum use, which could include recommended data definitions, structure, and formatting. The agency raises questions about the role of spectrum use data in band-specific proceedings and asks how it should frame the benefits and limitations of this data.

Comments on the NOI will be due in several months, after publication of the item in the Federal Register.