This article considers reliable and secure spectrum sensing (SS) based on federated learning (FL) in the cognitive radio (CR) environment. Motivation, architectures, and algorithms of FL in SSare discussed. Security and privacy threats on these algorithms are overviewed, along with possible countermeasures to such attacks. Some illustrative examples are also provided, with design recommendations for FL-based SS in future CRs.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Computer Networks and Communications
- Electrical and Electronic Engineering