A combined channel estimation method of time-frequency-selective fading channels in MIMO-OFDM systems was proposed.The time-varying channel was modeled as an autoregressive（AR） process and a low-dimensional Kalman filter based on pilots was used to estimate the AR parameters

and the LS（least square） algorithm was adopted to track the time-varying channel fading factors.The Kalman estimator explored the time-domain correlation of the channel

and a minimum mean square error（MMSE） combiner was used to modify the Kalman estimates.The proposed solution could reduce the complexity of the high-dimensional Kalman filter and track the channel both in frequency and time domains.The simulation results show that this method improves the LS estimates and has a comparable performance to the com-plex high-dimensional Kalman channel estimation method.