Reduced-Complexity Decision-Directed Channel Estimation in OFDM Systems with Transmit Diversity

Decision-directed channel estimation (DDCE) in orthogonal frequency-division multiplexing (OFDM) systems with transmit diversity suffers from high computation complexity in the simultaneous estimation of channel responses between the receiver and all the transmitters. Exploiting the correlation of the channel frequency response at adjacent subcarriers can help decouple the inter-antenna interference (IAI). This makes it possible to independently estimate each channel response, resulting in the reduction of computational complexity. However, existing IAI decoupling algorithms employ the least squares (LS) method for channel estimation from the IAI decoupled data. This approach is suboptimal because of the unequal variance of noise components in the decoupled data. Residual IAI, which arises from unequal channel gains at two adjacent subcarriers because of the channel frequency selectivity, occurs in the decoupled data, causing biases in the LS channel estimation. To improve the performance of reduced-complexity DDCE in transmit diversity OFDM systems, the proposed algorithm applies the best linear unbiased estimation method to independently estimate each channel response from IAI decoupled data. This approach is optimal regarding the unequal noise variances of the decoupled data. This study also exploits temporal channel correlation to remove residual IAI using the latest channel information of DDCE. Simulation results show that the proposed method can improve the performance of the reduced complexity algorithm, making it close to that of the joint estimation algorithm with significantly higher complexity.