Tenth Order Envelope Moment Based SNR Estimation Technique Using Newton-Raphson Method for AWGN and Rician Channels

Abstract

The Signal to noise ratio (SNR) estimation plays an important role in improving performance of adaptive wireless communication systems as well as software defined radio (SDR) or cognitive radio receivers in the future. SNR information can help in changing modulation schemes at the transmitter and also use different demodulation algorithms at the receiver to enhance performance and provide channel quality information required for handoff, power control and channel assignment algorithms. In this paper we propose and study a 10th order non data aided (NDA) moment based SNR estimation technique for QAM modulated signals over AWGN and Rician channel models. We use Newton-Raphson method for finding SNR from the resulting 5th order polynomial. This approach is more efficient than directly finding roots or use of look up tables considered in the literature and reduces computational and memory complexity considerably for even low order moment based methods proposed in the literature. We analyze the performance of the proposed estimator for QAM and PSK signals and compare with SNR estimation based on fourth order, sixth order and eighth order moments for AWGN as well as Rician channel models. Our simulation results show that the moment based SNR estimators work well for low SNR values that are of interest in communication system design.