| 摘要 |
PURPOSE: A channel equalizing method is provided to divide the entire section where an input signal is distributed into a plurality of sections to obtain an error for each of the divided sections and differently apply an error update rate to perform blind equalization. CONSTITUTION: The entire sections of signal constellation where input signals are distributed is divided into a plurality of sections according to probability that a decision directed error value is wrong, and an error is obtained for each of the divided sections. A tap factor of a filter is updated by differently applying an error update rate. When a channel-equalized signal is located in a region where probability that decision error values of real-number and imaginary-number parts are wrong is low, tap factors of the real-number and imaginary-number parts are updated with only decision directed errors.
|
