Myronchuk O. Methods of two-stage estimation of multipath channel parameters in communication systems with OFDM technology

PhD thesis in the field of knowledge 17 Electronic and Telecommunications in specialty 172 Telecommunications and Radio Engineering. – National Technical University of Ukraine «Igor Sikorsky Kyiv Polytechnic Institute», Kyiv, 2020. An actual scientific problem of developing of methods of two stage estimation of multipath channel parameters in communication systems with OFDM technology that uses known channel statistical parameters has been solved. The developed methods allow to increase the accuracy of channel frequency response estimation and to reduce the probability of error during the receiving of transmitted information. The first section shows the relevance of the problem of estimating the channel frequency response distortion, which is caused to multipath propagation of the signal and the Doppler effect. The principles of OFDM symbol formation and their structure was considered. An overview of pilot signal placement schemes in the structure of OFDM symbols and features of their application has been performed. Methods of channel frequency response estimation on subcarriers with pilot signals and methods of interpolation of frequency response values on subcarriers with data were considered. The comparative analysis of the considered methods on a model example by statistical modeling was carried out. The simulation results showed that the problem of developing methods for estimating the parameters of multipath channels with a long delay time and a significant influence of the Doppler effect is urgent. The research problem was formulated. In the second section the method of two-stage channel frequency response estimation using only currently received OFDM symbol has been developed. At the first stage, the values of the frequency response are filtered on the pilot subcarriers and extrapolated to the information subcarriers using the Kalman filtering algorithm in the forward and backward directions relative to the observation vector. In the second stage, the combining of parameters of a posteriori distributions on each subcarrier is performed. The method of two-stage channel frequency response estimation using estimates from previously received OFDM symbols has been developed. At the first stage, the values of the frequency response are filtered on the pilot subcarriers and extrapolated to the information subcarriers using the Kalman filtering algorithm in the forward, backward and time directions. In the second stage, the combining of parameters of a posteriori distributions on each subcarrier is performed. Using the statistical modeling the comparative analysis of the developed methods with known ones was carried out. In the third section the method of two-stage joint estimation of information symbol and channel frequency response using only currently received OFDM symbol has been developed. At the first stage, a recurrent calculation of the parameters of the a posteriori distributions of the values of the frequency response and the a posteriori probabilities of the information symbols in the forward and backward directions relative to the observation vector is performed. At the second stage the combining of parameters of a posteriori distributions and specification of a posteriori probabilities on each subcarrier is carried out. The method of two-stage joint estimation of information symbol and channel frequency response using estimates from previously received OFDM symbols has been developed. At the first stage, a recurrent calculation of the parameters of the a posteriori distributions of the values of the frequency response and the a posteriori probabilities of the information symbols in the forward, backward and time directions on each subcarrier is performed. At the second stage the combining of parameters of a posteriori distributions and specification of a posteriori probabilities on each subcarrier is carried out. Using the statistical modeling the comparative analysis of the developed methods with known ones was carried out. In the fourth section the analysis of the probabilistic characteristics of communication systems using the developed methods was performed. The calculation of computational costs for the implementation of the developed methods and comparison with the computational costs for the implementation of known methods was carried out. Based on the results of simulations and calculations, conclusions about the effectiveness of the developed methods and the possibility of their application in practice were made.