Tracking Effective Length of Channel Impulse Response for ML Estimator in OFDM Systems

This paper deals with the estimation of the channel impulse response (CIR) in orthogonal frequency division multiplexing (OFDM) systems. In particular, we focus on the maximum likelihood estimator (MLE). The MLE is simpler to implement as it needs no information on the channel statistics, but the channel estimation performance depends on the assumed effective length of channel impulse response (ELCIR), we propose a novel ML channel estimator which combine the ML estimation with a strategy of linear search. The proposed method introduces a few complexities, but, compared with the existing ML channel estimators, it is highly robust to variation of ELCIR. Its performance has been proved using theoretical analysis and simulations.