针对分布式多天线信道随机时变特征参数的获取问题,通过参数化建模方法建立信道时变参数的自回归模型,将由频率偏置和复信道衰落构成的强非线性观测方程在估计值处展开成泰勒级数进而线性化观测方程后,运用扩展卡尔曼滤波算法联合估计未知参数.仿真结果表明,该方法可在序贯的观测值下对信道时变参数进行联合估计和跟踪,能获得逼近克拉默—拉奥下界的估计精度.“,”In order to estimate random time-varying wireless channel characteristic parameters of distributed multi-input multi-output (MIMO) communication systems, autoregressive models of unknown parameters were established according to statistical models of MIMO channel. By linearization of observation equation composed of multiple frequency offset and complex channel fading at estimate values, it applied extended Kalman filter to jointly obtain the parameters. Simula-tion results show that the method can estimate and track time-varying parameters under sequential observation. It can ob-tain the estimate accuracy close to Cramer-Rao lower bound.