A Robust Signal Classification Scheme for Cognitive Radio

TL;DR

This paper introduces a robust, feature-based signal classification scheme for cognitive radio that effectively mitigates noise uncertainty and is validated through both simulations and real-world experiments.

Contribution

It proposes a novel dimension cancelation (DIC) method to enhance feature-based detection robustness against noise uncertainty in spectrum sensing.

Findings

Simulation results show high classification accuracy.

Experimental tests confirm robustness in real-world conditions.

Scheme effectively mitigates noise uncertainty issues.

Abstract

This paper presents a robust signal classification scheme for achieving comprehensive spectrum sensing of multiple coexisting wireless systems. It is built upon a group of feature-based signal detection algorithms enhanced by the proposed dimension cancelation (DIC) method for mitigating the noise uncertainty problem. The classification scheme is implemented on our testbed consisting real-world wireless devices. The simulation and experimental performances agree with each other well and shows the effectiveness and robustness of the proposed scheme.