The Research On Channel Estimation Algorithm Of NGB-W

NGB is Next-Generation Broadcast Network, that in conformity with the nationalconditions, convergence of three networks, combination of wired and wireless, andoverall network. It is based on the achievements of the digital cable television andmobile multimedia broadcasting, which in accordance with the requirements of thestate. NGB-W is an important part of the NGB. The NGB-W system adoptsMIMO-OFDM technique. It improved the ability of spectrum utilization efficiencyand the resistance to multipath fading. This thesis main studies the channel estimationalgorithm of NGB-W system.First, this thesis briefly introduced the development of broadcast televisionnetworks technology at home and abroad. Introduced the aim and meaning ofresearching NGB-W, and explain channel estimation had played an important role inwireless broadcast networks system. The MIMO-OFDM, coding, interleaving andconstellation mapping technology were introduced, these technology had been used inNGB-W system. The wireless channel propagation characteristic had simpleintroduced in this thesis, and at the same time, analyzed the influence of the systemcaused by fading.Next, this thesis introduced the method to design reasonable pilot pattern andobtain channel response through the less pilots. The channel estimation algorithm,such as LS, MMSE and LMMSE algorithm were indicated in this thesis. Besides, italso indicated constant interpolation, linear interpolation, Gaussian interpolation,cubic interpolation, Wiener filtering interpolation and based on DFT interpolationalgorithm. Next, it introduced ZF and MMSE this two equalization algorithm.LMMSE and Wiener filtering channel estimation algorithm to focus the analysis on.In this thesis, TU6channel model has been used to as simulation environment.LS, LMMSE and Wiener filtering estimation were performed. MSE and BERperformance evaluation were achieved. This thesis aimed at reducing the huge amountof computation of the correlation coefficient matrix when adopted Wiener filtering,it had adopted fixed coefficient Wiener Filtering estimation, and optimized thealgorithm on the base of pilot structure of system. Then the simulation result wasperformed and got a better performance. The fixed coefficient Wiener Filtering estimation algorithm has smaller losses in performance, but greatly reduced the relatedcalculation. In addition to, LS estimate of the frequency domain linear interpolation invariety of pilot pattern were performed. From the simulation results we can see that,from the pilot pattern PP1to PP5, channel estimation performance become worse，andit could give the pilot structure optimization to provide certain reference from thesimulation results. The influence of channel estimation algorithm performance causedby Doppler shift which comes from the movement were also performed in this thesis.