Iterative BEM channel estimation and BP detection for ICI cancellation in DVB systems with very high mobility

Modern terrestrial television broadcasting, which is based on orthogonal frequency division multiplexing (OFDM) transmission, suffers from inter-carrier interference (ICI) in time-varying channels, due to the loss of orthogonality among subcarriers. This ICI term, which cannot be neglected in middle and high mobility scenarios, can even offer increased diversity if it is correctly processed by the detector, which requires a good knowledge of the complete channel, including ICI. In recent years, BEM has been successfully employed to model and estimate time-varying channels using clustered frequency-domain pilots. In this paper, we propose to combine a complex-exponential CE-BEM channel estimator with a belief-propagation based ICI-aware detector in an iterative (turbo) receiver architecture. The soft information transfer among the different receiver blocks, the iteration patterns and the design of optimized pilot clusters are analyzed for high-mobility scenarios. Simulation results show the very good performance of the proposed novel joint algorithm that not only cancels ICI, but also can benefit from the diversity it introduces, providing specially good results under high mobility scenarios.