Hybrid Spectrum Sensing Architecture for Cognitive Radio Equipment

Spectrum sensing is the key function in implementing cognitive radio, which enables secondary users to identify and utilize vacant spectrum resource allocated to primary users. Recent studies have proposed four major sensing methods, including matched filter, energy, feature, and eigenvalue-based detectors. However, there are some drawbacks along with them. In this paper, we propose a hybrid architecture, associating energy and cyclostationary detectors, for spectrum sensing that improves the ability of conventional energy detector to detect the primary user in the presence of noise uncertainty. In a constant noise environment, the performance of the proposed detector approaches that of an ideal radiometer.