The authors propose low complexity versions of a linear minimum mean square error (LMMSE) channel estimator for orthogonal frequency division multiplexing (OFDM) systems. They partition the channel autocorrelation matrix in the LMMSE estimator into small sub-matrices. The size of sub-matrix is determined by channel coherence bandwidth, a range of highly correlated channel frequency responses. Since the proposed estimator uses small sub-matrices instead of the whole channel auto-correlation matrix, the complexity is significantly reduced in spite of slight performance degradation. The performance of mean square error (MSE) has been evaluated in an exponential decay channel model through computer simulation to verify the validities of proposed schemes.