A Very Low Complexity LMMSE Channel Estimation Technique for OFDM Systems

In this paper, we propose a very low complexity Linear Minimum Mean Square Error (LMMSE) channel estimation technique for Orthogonal Frequency Division Multiplexing (OFDM) systems. The conventional LMMSE channel estimation technique suffers from a high computational complexity with an order of O(N_P^3) operations (N_P represents the total number pilot in one OFDM symbol) due to an inversion matrix operation. By exploiting the structure of the channel autocorrelation matrix, we propose a very low complexity LMMSE channel estimator based on a FFT operation, a circle shift operation and a fast Toeplitz matrix-vector multiplication. The complexity of the proposed algorithm is reduced to O(N_P logN_P) operations. Simulation results show that the proposed technique has the same performances as well as the classical LMMSE in terms of Bit Error Rate (BER) and Mean Square Error (MSE).