Abstract: SNR(Signal-to-Noise Ratio) is an important parameter in modern communication signal processing, and many algorithms need this information to obtain optimal performance. A new blind SNR estimation algorithm is proposed for signal-to-noise ratio (SNR) estimation over single-input multiple-output (SIMO) system in this paper. The new SIMO SNR estimation technique is based on singular-value decomposition (SVD) of received signal covariance matrix, by computing the maximum eigenvalue of the matrix to estimate SNR. The algorithm can estimate the SNR of digital modulation signals commonly used in additional white Gaussian noise(AWGN) channels and multipath channels without prior information of the modulation type. The simulation results show that this algorithm offers excellent performance. Compared with the SVD-based method in single-input single-output(SISO) system, the proposed algorithm can enhance the estimating accuracy without computing the dimensions of signal and noise space, and achieve estimations with a small computational complexity.