Signal Wiener Filter, The Self-tuning Information Fusion

For the multisensor system with unknown model parameters and noise statistics, by the modern time series analysis methed, based on on-line identification of the autoregressive moving average(ARMA) innovation model parameters by the lease squares methed the unknown model parameters and noise variances can be estimated. Under the linear minimum variance optimal information fusion criterion weighted by scalars, a self-turing distribuced information fusion white noise deconvolution filter is presenced, two self-turing distribuced information fusion Wiener filter are presenced for ARMA signals, and a self-turing distribuced information fusion Wiener deconvomation filter is presenced for AR signal. Based on optimal weighted measurement fusion methed, a self-turing weighted measurement fusion white noise deconvolution filter is presenced, a self-turing Weighted measurement fusion Wiener filter is presenced for ARMA signal, and a self-turing Weighted measurement fusion Wiener deconvomation filter is presenced for AR signal. Many simulation examples show their effectiveness. The simulation results show that the accuracy of self-turing fusion is higer than that of each local filters,and they have asympotic optimality, i.e if the parameter estimation of the ARMA innovation model is consistent, then they will converge to the optimal fusion with known model parameter and noise statistics.