Multi-stage information fusion identification method for multisensor ARMA signals with white measurement noises

For the single channel autoregressive moving average (ARMA) signals with multisensor and white measurement noises, unknown model parameters and unknown noise variances, a multi-stage information identification method is presented. It consists of the recursive instrument variable (RIV) algorithm-based information fusion estimator of the autoregressive (AR) parameters, the correlation method-based information fusion noise variance estimators, and the correlation function-based information fusion moving average (MA) parameter estimator. All information fusion estimators are obtained by taking the average of the local estimators. They have the strong consistency. A simulation example shows its effectiveness.