Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 99 === This thesis presents a new method for designing fixed-point FIR filters. In the proposed method, both magnitude response and hardware cost are taken into consideration in the design process. The idea of the common sub-expression sharing method is applied by determining a basic set and expanding this basic set gradually. In addition, to deal with long filters, some strategies that aim to speed up the program are also employed. The experimental results of several benchmark examples show that by applying the proposed method, we can obtain the second best results, in an acceptable run time than the best results of published papers.
|