Channel prediction by Doppler frequency estimation with simplified AR modeling

A channel prediction method using the estimation of Doppler spectrum is proposed and investigated to predict channel state information in time varying channel. Because the Doppler frequency is one of main features for time varying channel characteristics, the autoregressive model for the Doppler spectrum could be employed to get reliable channel prediction coefficients. For the IIR approximation of Doppler spectrum, we propose cascade filter of digital Biquad and low pass filter with simple modification. In the cascade filter, we need only two parameters, and we can get higher order channel prediction by cascading same filter as many as we want.