ABSTRACT
This chapter argues that the answer is no: radio spectrum management cannot be “outsourced” to software. Global spectrum management comprises a complex matrix of primary law, regulations, international treaties, best practices, and property rights. The telecoms industry and Artificial Intelligence (AI) are already intertwined. Much of the data used in AI are gathered by Internet of Things networks, while aspects of modern and future telecoms networks use AI predictive analytics capabilities to manage components in their networks. Machine Learning implementations of AI can also help radios better perform various functions, such as spectrum sensing, Radio Access Technology classification, signal classification, Medium Access Control protocol identification, Power Allocation, and Attack detection. AI promises to provide dynamic spectrum assignments that respond to real demand for radio spectrum at any given moment. Policymakers may consider rules requiring that devices transmit information about their own identities, or that devices maintain logs of their activities that can be checked after an interference event.
