Secure Federated Learning for Cognitive Radio Sensing
Małgorzata Wasilewska, Hanna Bogucka, H. Vincent Poor
Abstract
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.
Topics & Concepts
Computer scienceCognitive radioComputer securityComputer networkTelecommunicationsWirelessDistributed Sensor Networks and Detection AlgorithmsCognitive Radio Networks and Spectrum SensingWireless Communication Security Techniques