On sequential blind channel estimation for time-varying OFDM system

A new algorithm is proposed for the sequential blind channel estimation of the Orthogonal Frequency Division Multiplexing (OFDM) systems. Different from other blind algorithms which carry out estimation for a fixed set of subchannels or time-slots, the new algorithm based on the Expectation-Maximization (EM) method is offered to obtain the Maximum-likelihood (ML) estimation of the time-domain channel impulse response (CIR) sequentially. We also propose a novel recursive approach to get the parameter value. This recursive approach not only reduces computing complexity but also avoids computing instability. It's further shown that this algorithm can be simplified mathematically by combining iteration and recursion. Simulations illustrate these algorithms provide better performance than other algorithms.