Improving spectrum sensing performance in cognitive radio networks using MIMO entropy-based detection

The rapid growth in the number of wireless devices in recent years has intensified pressure on the radio spectrum, making static spectrum allocation increasingly inefficient. Despite fixed allocation policies, numerous studies have shown that large portions of licensed spectrum remain underutilized. Cognitive Radio (CR) technology addresses this challenge by enabling dynamic spectrum access, allowing the detection and utilization of unused frequency bands. Within CR systems, spectrum sensing is essential for identifying available spectrum holes. Energy Detection (ED) is widely adopted due to its simplicity, low computational cost, and capability to sense various primary user (PU) signals. However, its performance significantly degrades under low Signal-to-Noise Ratio (SNR) conditions due to noise uncertainty. To mitigate this issue, this study employs MIMO-based ED and MIMO-based entropy detection techniques. Among these, different entropy measures are evaluated, and the results indicate that Rényi entropy detection provides the most substantial improvement in overcoming the SNR wall, outperforming Tsallis, Shannon, and Kapur’s entropy methods. Consequently, this thesis proposes MIMO entropy-based spectrum sensing to enhance single-stage detection performance, demonstrating that the MIMO Rényi entropy approach achieves superior performance compared to other entropy-based MIMO detectors.