Two-Stage Method for Joint Estimation of Information Symbols and Channel Frequency Response in OFDM Communication Systems

Abstract

The optimal and quasi-optimal algorithms of two-stage joint estimation of information symbols and frequency response (FR) of multibeam channels in OFDM-based communication systems have been synthesized. They involve the calculation of a posteriori probability density of estimated processes. At the first stage, the recurrent calculation of joint a posteriori distributions of information symbols and the frequency response is performed from two sides of the measurement vector. At the second stage, a posteriori distributions are combined on each of subcarriers that are obtained as a result of the first stage estimation. Based on the obtained a posteriori distributions, estimates of the information symbol and channel FR are determined by the criteria of a posteriori probability maximum and the average mean squared error minimum, respectively. The performance of synthesized algorithms similar to the well-known MMSE algorithm involves the need of knowledge of statistical properties of communication channel. The device implementing the optimal algorithm is multichannel, each channel of which is matched with the corresponding value of symbol from the modulation constellation. The quasi-optimal algorithm was obtained by using the Gaussian approximation of a posteriori probability density. It should be noted that this algorithm preserves its multichannel structure. The approbation of the presented algorithm was conducted using the computer-aided statistical simulation for different parameters of multibeam communication channels and the comparison of the obtained results with the results obtained by the MMSE and LS algorithms. The simulation results showed that the error probability decreases by about 2 times with a maximum signal propagation delay of 100 μs.