Blind propagation channel estimation with enhanced auto-deconvolution
Rabah Maoudj, Ali Dziri, Michel Terré · 2016
This paper deals with blind propagation channel estimation that should be a key feature for future agile waveforms. The proposed approach is based on high order statistics, inverse Fourier Transform and auto-deconvolution. The paper can be seen as a follow up to a previous work devoted to this subject [10]. It introduces an important improvement in the auto-deconvolution step based on a whitening algorithm able to equalize additive noise corrupting high order statistical moments estimation. Simulations results highlight the significance of this new stage, showing its impact, not only on the mean square error, but also on the bit error rate for 4 to 64-QAM modulations.