Radio Frequency-Enhanced Multi-Factor IoT Device Authentication Via Swarm Learning

With the increased popularity of Internet of things (IoT) devices, security issues have notably risen in recent times. Typically, wireless IoT applications are vulnerable to impersonation attacks by malicious entities. This paper proposes a lightweight multi-factor authentication mechanism boosted by radio frequency fingerprinting (RFF) to physically identify IoT devices. A novel application of swarm learning (SL) is utilized to develop the authentication model and enable distributed authentication. This approach maintains privacy and is resilient against faults when processing RFF data from various sources. The device-side multi-factor authentication is lightweight and has been validated through a formal security model. Experimental results indicate that the proposed scheme achieves the highest authentication success rate and the lowest computational cost on the device side compared to other authentication methods, which also validated its effectiveness in defending against impersonation and poisoning attacks.