Maximum a posteriori semi-blind channel estimation for OFDM systems operating on highly frequency and time selective channels

We propose a semi-blind block-by-block channel estimation algorithm for OFDM systems operating in highly frequency and time selective channels. This algorithm, based on the Expectation-Maximization algorithm, performs an iterative maximum a posteriori estimation of the channel. It can incorporate the coded structure of each coded block in a turbo-processing fashion to enhance channel estimation quality Moreover; it call use the maximum a posteriori training-based channel estimation algorithm for its optimum initialization. The performance of this channel estimation algorithm is evaluated through simulation results, with and without turbo-processing and optimum initialization. It is also compared to two benchmarks based on least-square error channel estimation. Our algorithm can be potentially applied for the efficient estimation of the propagation channel in DRM, DVB-T, HIPER-LAN/2 and IEEE 802.11a systems.