Channel Estimation In MSE-OFDM System

OFDM has found wide applications due to its capability in reducing the multi-path effect, but it has the drawbacks of high Peak-to-Average Power Ratio (PAPR) and being sensitive to frequency offset. A flexible transmission technique termed Multiple-Symbol Encapsulated OFDM (MSE-OFDM) with improved bandwidth efficiency and robustness to various interferences was proposed. Study shows that it has a better performance than the conventional OFDM and single-carrier systems in frequency-selective fading channels.This thesis researches on the channel estimation algorithm in MSE-OFDM systems. Based on the principle and the structure of MSE-OFDM system, the channel estimation algorithms with frequency domain and time domain pilot are studied, and a better training sequence in time domain pilot estimation algorithm is proposed. The author's main contributions are as follows:1. In the frequency domain pilot channel estimation, this thesis studies the pilot subcarrier and data subcarrier channel estimation algorithms. The author studies and compares the pilot subcarrier estimation methods of LS, MMSE and SVD. In the data subcarrier estimation methods, the first-order interpolation and FFT interpolation methods are studied and compared.2. In the time domain pilot channel estimation, this thesis researches an algorithm based on the maximum likelihood principle. By minimizing the mean square error of channel estimation, the optimal training sequence suitable for the channel estimation method is obtained.3. The training sequence is further optimized using genetic algorithm and a training sequence with constant PAPR and flatter in frequency domain is obtained. Simulation results show that the proposed channel estimation method is feasible for improving the performance of MSE-OFDM with a training sequence flat in time and frequency domain.