Frequency Flat Unknown Fading Channels

When perfect channel state information (CSI) is available at a receiver, the maximum Doppler diversity gains from frequency flat time selective (FFTS) fading channels can be achieved with a pair of judiciously designed precoding and grouping scheme. When CSI is not available at the receiver, modelling FFTS fading channels with basis expansion models (BEM) enables channel estimation without knowledge of channel statistics. However, the major drawback with BEM based channel estimators (BEM-CE) is that bit error rate (BER) performance of the receiver suffers from high error floor due to inaccurate channel modelling. Till this date, an effective transmitter-based solution to improve the BER of receivers with BEM-CE has not been proposed yet. In this paper, we propose a novel grouping and precoding scheme to minimize the pairwise error probability (PEP) of a receiver with BEM-CE. Both experimental and theoretical results show significant improvement in the BER performance of the receiver with the PEP-minimizing grouping and precoding scheme.