| 摘要 |
A method and adaptive filter for processing a sequence of input data for an audio system may have the input data provided to first and second predictive filters, and the data may be delayed in time before or after the second filter. One of the predictive filters may be backward-looking while the other may be forward-looking, and the data filtered by the two predictive filters may be added to yield the output data. The input data may be delayed by a second delay, which may be less than the first delay, and the resulting delayed data may be used in determining a prediction error signal by subtracting the data filtered by the first predictive filter from the input data delayed by the second delay and by providing the resulting data to the first predictive filter. Alternatively, the prediction error signal may be determined by subtracting the output data from the input data delayed by the second delay. The prediction error signal may be provided to the two predictive filters. The prediction error signal may be provided to the first predictive filter for its adaptive control with tap values, and complex conjugated tap values may be provided to the second predictive filter for its adaptive control. The input data delayed by the second delay may be subjected to an indirect center tap control which receives tap values or complex-conjugated tap values of the first predictive filter as the control signal, and the output values of the center tap control may be added to the data filtered by the two predictive filters. The data from a received signal with multiple signal components of static or quasi-static or slowly varying frequency may be used as the input data, in particular, the input data of a digital OFDM television signal.
|
