Signal Detection and Spectrum Sensing Using Random Matrix Theory in Massive MIMO Systems

Spectrum sensing plays a key role in modern communication systems where higher congested bands require more intelligent radio access. Additionally, next generation wireless, such as 5G, will use a large number of antennas and increased carrier frequency ranges. Unfortunately, due to high frequency carriers resulting in greater path loss, in low signal-noise-ratio (SNR) environments many spectrum sensing approaches do not work effectively. A random matrix theory (RMT) generalized likelihood ratio test (RGLRT) algorithm is introduced in this paper for massive multiple-input multiple-output (MIMO) antenna systems. We successfully demonstrate the eigenvalue distribution of the covariance matrix of the MIMO system signal is the same as a spiked central Wishart matrix. The spike dimension of this central Wishart matrix is associated to the number of users. Using this result, the highly recognized (linear) generalized likelihood ratio test (GLRT) is implemented for hypothesis testing. Compared with other well-known spectrum sensing methods, this new algorithm gives impressive detection performance in low SNR environments when the number of antennas is large. Simulations further confirm the theoretical results.