A Hilbert-Wavelet Fusion for Enhanced Wideband Spectrum Sensing in Cognitive Radio

Abstract

The rapid growth of wireless technology has escalated demand for the finite radio spectrum. Cognitive Radio addresses this by enabling dynamic access to underutilized bands, yet its effectiveness depends on accurate spectrum sensing. Current sensing methods often suffer from poor performance in noisy environments and low-signal conditions, creating a critical reliability gap. This study presents an enhanced sensing technique that integrates the Hilbert Transform with Discrete Wavelet Packet Transform. The hybrid design sharpens signal analysis across sub-bands, and an adaptive threshold mechanism improves vacancy detection. The system was rigorously evaluated through simulation. Key results demonstrate a significant performance improvement. The proposed technique achieves, on average, a 77% better signal representation across decomposed channels compared to conventional wavelet sensing. Furthermore, Receiver Operating Characteristic (ROC) curves confirm a higher probability of detection at signal-to-noise ratios of -5dB, 0dB, and 5dB. These metrics substantiate that the fused method offers superior detection capability and reliability. The work provides a validated approach for making cognitive radio networks more robust and spectrum utilization more efficient.