A Constrained GLRT Framework for Blind Spectrum Sensing in OFDM Based Cognitive Radio Systems

Cognitive radio (CR) technology has emerged as a promising solution to address spectrum scarcity by enabling opportunistic access to licensed frequency bands. This paper presents novel blind spectrum sensing algorithms for orthogonal frequency division multiplexing (OFDM) based cognitive radio systems using the constrained generalized likelihood ratio test (C-GLRT) framework. We demonstrate that the existing cyclic prefix correlation coefficient (CPCC) based detection algorithm is a special case of C-GLRT in additive white Gaussian noise (AWGN) channels. However, the performance of CPCC based sensing degrades significantly in multipath fading environments, which are typical for OFDM applications. To address this limitation, we propose a multipath correlation coefficient (MPCC) based C-GLRT algorithm that exploits the inherent structure of the OFDM channel matrix. Furthermore, we develop a combined CPCC-MPCC detection algorithm that leverages both cyclic prefix and multipath correlations to achieve superior detection performance. The proposed algorithms operate without requiring prior knowledge of the primary user signal, channel state information, or noise variance, making them truly blind detection schemes. Simulation results demonstrate that the proposed MPCC based C-GLRT and combined algorithms significantly outperform conventional energy detection and CPCC based methods, particularly in low signal-to-noise ratio (SNR) regimes and rich multipath environments. The combined algorithm achieves near-optimal detection performance while maintaining computational efficiency suitable for practical CR implementations.