Time-Domain Least Square Channel Estimation for Polarization-Division-Multiplexed CO-OFDM/OQAM Systems

Suppressing the intrinsic imaginary interference (IMI) effect induced by the multiple-path fading channel is the key technology for maintaining the good performance of the orthogonal frequency-division multiplexing offset-quadrature amplitude modulation (OFDM/OQAM) systems. Recently, the first theoretical discussion on the IMI effect and the corresponding frequency-domain channel estimation method for polarization-division-multiplexed (PDM) coherent optical OFDM/OQAM (CO-OFDM/OQAM) system has been studied. The full-loaded and the half-loaded channel estimation methods have been proposed to mitigate the IMI effect. However, for these frequency-domain methods, the condition that the symbol interval is much longer than the maximum channel delay spread has to be satisfied. When the transmission distance is long, the frequency-domain residual errors induced by the chromatic dispersion (CD) and polarization mode dispersion (PMD) reduce the channel estimation accuracy evidently. In this paper, we systematically discuss the time-domain channel transmission model for PDM CO-OFDM/OQAM systems. With the analysis of the distribution of the received additive noise, we propose the time-domain least square (TDLS) channel estimation method for PDM CO-OFDM/OQAM systems. Compared with the frequency-domain methods, the TDLS method promotes the system robustness against both the IMI effect and the additive noise significantly. The computational complexities and the transmission performance have been compared for both the TDLS and the frequency-domain full-loaded methods. The theoretical analysis is validated by numerical Monte Carlo simulations of the PDM CO-OFDM/OQAM system.