Linear Prediction-Based Approach to SNR Estimation in AWGN Channel

A general technique for signal-to-noise (SNR) estimation of digital communication signals corrupted by additive white Gaussian noise (AWGN) is described. The proposed SNR estimator views the problem in terms of interpolation of the power of the noise-free data signal based on the second order characteristics of corrupted signal with additive noise. A linear predictor with coefficients estimated based on the least-squares solution of linear prediction errors combined in forward and backward directions is used to predict the variance of the additive noise. The performance of the proposed technique is investigated and compared with other SNR estimators by computer simulation of baseband binary phase-shift keying (PSK) signals in real additive white Gaussian noise (AWGN) and baseband 8-PSK signals in complex AWGN