An improved channel estimation based on generalized DFT for OFDM system

The traditional channel estimation based on Discrete Fourier Transform (DFT) for Orthogonal Frequency Division Multiplexing (OFDM) system has significantly outperformed least square (LS) or minimum mean-square error (MMSE) channel estimation by eliminating the impact of noise beyond the maximum channel delay. However, its performance is degraded due to the presence of set correlations among the basis functions. In order to overcome the problem, Generalized Discrete Fourier Transform (GDFT) with nonlinear phase functions which is a generalized theoretical framework for DFT has been introduced to estimate channel in this paper. GDFT can be designed optimally for phase functions based on a variety of correlation metrics to reduce inter-symbol interference (ISI) and inter-carrier interference (ICI) that determine the accuracy of the channel estimators. Simulation results show that GDFT improves the correlation performance dramatically over DFT and channel estimation based on GDFT has better bit-error-rate (BER) performance than that based on DFT.