Artificial Channel Aided LMMSE Estimation for Time-Frequency Selective Channels in OFDM Context
Résumé
This paper proposes a linear minimum mean square error-based (LMMSE) channel estimation method, which allows avoiding the necessary knowledge of the channel covariance matrix or its estimation. To do so, a perfectly tunable filter acting like an artificial channel is added at the receiver side.We show that an LMMSE estimation of the sum of this artificial channel and the physical channel only needs the covariance matrix of the artificial channel, and the channel estimation is finally obtained by subtracting the frequency coefficients of the added filter. We call this method artificial channel aided-LMMSE (ACA-LMMSE). Theoretical developments and simulations prove that its performance is close to theoretical LMMSE, and we show that this method reduces the computational complexity, compared to usual LMMSE, due to the covariance matrix used for ACA-LMMSE is computed only once throughout the transmission duration. We put the conditions on the artificial channel parameters to get the expected mask effect. Simulations display the performance of the proposed method, in terms of MMSE and bit error rate (BER). Indeed, the difference of BER between our method and the theoretical LMMSE is less than 2 dB. becoming an important topic especially when it comes to multistandard designs. This paper capitalizes on the Common Operator technique to present new common structures for the FFT and FEC decoding algorithms. A key benefit of exhibiting common operators is the regular architecture it brings when implemented in a Common Operator Bank (COB). This regularity makes the architecture open to future function mapping and adapted to accommodated silicon technology variability through dependable design.