Code-Aided Iterative SNR Estimator for M-APSK Signals Based on Expectation Maximization Algorithm

Abstract

A code-aided (CA) iterative signal-to-noise ratio (SNR) estimator based on Expectation Maximization (EM) algorithm is proposed for M-ary amplitude phase shift keying (APSK) signals. The estimation algorithm utilizes a posteriori probabilities of coded bits obtained from channel decoder to improve estimation precision at low SNRs. Furthermore, Cramer-Rao bound (CRB) of the proposed CA iterative SNR estimator for M-APSK is derived and simulated numerically. Compared with the non-data-aided (NDA) EM-based estimator and momentsbased estimators for M-APSK signals, computer simulation results show that the proposed estimator exploiting a posteriori information has more excellent performance, especially at low SNRs. It is also demonstrated that the performances of the proposed CA SNR estimator for 16-and 32-APSK signals are very close to the derived CRB.