A Kalman based SISO equalizer for HF ionospheric channel

Communication over HF-ionospheric channel requires the use of a performant receiver based on turbo equalization scheme. In this receiver, channel estimation can be improved for each iteration using soft information fed back from the decoder. We propose a turbo equalizer based on a soft-input Kalman estimator that can operate without channel impulse response knowledge. Performance analysis of the channel estimator is tested using four paths channel for two Doppler spread values and records of realistic channel. Simulation results show the convergence of the algorithm versus the frequency dispersion of the paths and compare the BER of the proposed Kalman based turbo equalizer with that of the conventional turbo equalizer. Results show that the proposed equalizer provide a significant gains in BER performance and convergence without the need for training data.