Subcarrier SNR ML estimators and Cramér-Rao bounds in multicarrier transmission

In this paper, considering a multicarrier transmissions system, we propose two techniques of maximum-likelihood (ML) subcarrier signal-to-noise ratio (SNR) estimation, in both data-aided (DA) and non-data-aided (NDA) schemes, and derive the corresponding Crame¿r-Rao lower bounds (CRLBs). The channel gains and phases are assumed to be constant over the observation window and the received signal is assumed to be corrupted by additive white Gaussian noise (AWGN). The proposed SNR estimators both exploit the mutual information between the different subcarriers and reach the corresponding CRLBs, as shown by Monte-Carlo simulations.