A High Accurate and Low Bias SNR Estimator: Algorithm and Implementation

Abstract

Signal to noise ratio (SNR) estimators are required for many radio engineering applications. In this paper, a SNR estimator based on the first and second order moments is derived and examined for constant envelop modulations over additive white Gaussian noise (AWGN) channel. In the first, a modification method is proposed to reduce the bias of conventional first and second order moments based SNR estimator. Then, in order to reduce hardware implementation complicity, multi segments of cubic polynomial are adopted to approximate the expression of the proposed SNR estimator which doesn’t have a close form solution. The approximate expression is very precise in the SNR range -10dB ~ 20 dB as illustrated by numerical simulation. Besides, practical hardware circuit is proposed for FPGA implementation. Simulation results show that the proposed SNR estimator has the lowest normalized bias and variance when compared with two other classic estimators.