We investigate the problem of spectrum sensing in cognitive radios (CRs) when the receivers are equipped with a large array of antennas. We propose and derive three detectors based on the concept of linear spectral statistics (LSS) in the field of random matrix theory (RMT). These detectors correspond to the generalized likelihood ratio (GLR), Frobenius norm, and Rao tests employed in conventional multiple antenna spectrum sensing (MASS). Subsequently, we compute the Gaussian distribution of the proposed detectors under the noise-only hypothesis, leveraging the central limit theorem (CLT) applied to high-dimensional random matrices. We evaluate the performance of the proposed detectors and analyze the impact of the number of antennas and samples on their efficacy. Furthermore, we assess the accuracy of the theoretical results by comparing them with simulation outcomes. The simulation results provide evidence that the proposed detectors exhibit efficient performance in wireless networks featuring large array antennas. These detectors find practical applications in diverse domains, including massive MIMO wireless communications, radar systems, and astronomical applications.