SNR Estimation Technique of the AWGN Channel by Second and Fourth-Order Moments (fufrMt)

Abstract

The goal is to Jind the "best" estimate of the SI{R in a digital receiver with the lea.st cost. Generalll,, SNR estimates are generated by averaging obser,*able properties o.f the received signal over a number o.f symbols. The performances of various signal-to-noise ratio (SNR) estimation techniques reported in the literatttre are compared to identify the "best" estimator. The SlrlR estimators are investigated b.v the computer simulation of baseband binary phase shift keying (PSK) signals in real additive white Gaussian noise (A't4rcl'{) and baseband M-ary PSK signals in complex Additive white Gaussian noise (AWGI'|). Some known estimator structltres are modified to perform better on the channel of interest. lt investigates, therefore, both types of estimators and quantifies by example the improvement in performance achievable by using known data rather than error-corruptecl recovered data. Estimator structures for cotnplex channels are examined.