Channel estimation based on continuous pilot channel with regression in LAS-CDMA

For downlink transmission, known pilot symbols are usually used to evaluate the channel characters through the common pilot channel. Contrasting to traditional CDMA, LS spreading codes in LAS-CDMA (large area code division multiple access) have IFW (interference free window). Thus when pilot symbols are used to estimate the channel coefficients. AWGN (additive white Gaussian noise) is the main obstacle to the accuracy. To suppress the noise degradation, in this paper, a new channel estimation method for LAS-CDMA system is introduced. In the proposed method, the channel coefficients obtained through common pilot channel are first fed to a regression process and these updated coefficients are then for the corresponding data symbol recovery. Linear and nonlinear regression schemes are both considered according to vehicle speeds and regression process lengths over the ITU-VA channel. Simulation results prove that this method can improve the channel estimation accuracy and system BER (bit error rate) performance greatly.