Optimal feedforward estimation of frequency-selective fading, radio channels using statistical channel information

The optimal feedforward channel estimator relying only on known training symbols is systematically derived. Statistical channel information is assumed to be available. The resulting estimator is shown to decompose into the two subtasks of maximum likelihood (ML) acquisition and subsequent Wiener filtering. When using optimal training sequences, the corresponding channel estimator becomes very efficient and flexible. A design example of a near-optimal HF channel estimator is presented and its performance was verified by simulation. Bit error rate simulation results of an entire near-optimal maximum likelihood sequence estimation (MLSE) receiver employing feedforward synchronization demonstrate the superiority and robustness of Wiener channel estimation