Table 2.

Federated learning use cases in IoT.

Ref.	IoT Use Cases	Specification of FL	Limitations
[35]	Attack defense	An FL-based attack defense network was proposed to secure industrial IoT networks	no consideration is given for the impact of latency in communication.
[39]	Attack detection	A FL model to detect security attacks in IoT	The problem of data privacy has been overlooked
[40]	Attack detection	FL-based attack detection in industry 4.0	No research work has been carried out on scalability issues
[41]	Intrusion detection	An FL-based intrusion detection model for IoT	Its performance is not considered
[42]	Intrusion detection	Intrusion detection system using FL in IoT	The performance is not validated by comparing with ML and DL approaches
[41]	Malware detection	Malware detection in Android applications using FL	The confluence of training process is overlooked
[43]	Intrusion detection	A review of FL techniques for detecting intrusion is considered	Coordination between different IoT devices is a major problem
[44]	Data breaching	FL-based identification and prevention of data breaches in industrial IoT	Larger datasets need to be tested
[45]	Malware detection in IoT devices	FL-based security enhancement in IoT	Energy performance and learning ability is overlooked