Abstract
In this paper, we propose a blind binary low-density parity-check (LDPC) encoder identification scheme for M-quadrature amplitude modulation (M-QAM) signals. The expectation-maximization (EM) algorithm is developed to estimate the unknown signal amplitude, noise variance, and phase offset for M-QAM signals. The a posteriori probabilities (APPs) of the coded bits are obtained from the APPs of the transmitted symbols according to the M-QAM mapper. Monte Carlo simulation results demonstrate the effectiveness of our proposed new blind binary LDPC encoder identification scheme for different modulation orders. The average iteration number needed for the EM algorithm to converge is also investigated for different modulation orders.