On Channel Estimation in MB-OFDM UWB Systems with Time Varying Dispersive Fading Channel

New channel estimation technique is proposed for multiband orthogonal frequecny division multiplexing (MB-OFDM) ultra wideband (UWB) systems in multipath time varying wireless channels. Two-stages approch has been used to achieve this purpose. In first stage, Winner-Hopf filtration has been employed for the interpolation of unknown channel state information (CSI) using comb-type known pilots. In second stage, interpolated channel statistics are then modeled as autoregressive (AR) process and fed into kalman filter. Moreover, inorder to suppress intercarrier interference (ICI), an ICI mitigation filter does joinly work with kalman filter. A mathematical framework is given for the ralization of our proposed system. Link level simulation (LLS) urgs that this new tecnique shows exact channel tarcking and provides better symbol error rate (SER) performance.