Synthesis of Time-Varying IIR Filters with Finite Word-Length Using Integer Programming

• Main Authors: Wang Shu-hung, 王樹宏
• Other Authors: Tzong-Yeu Leou
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/21584331699894871031

碩士 === 國立臺灣科技大學 === 電子工程技術研究所 === 86 === This thesis addresses the problem of designing finite word-length infinite impulse response (IIR) digital filters, which can be formulated as a discrete optimization problem. The design technique utilizes the local search algorithm of the integer programming which defines a larger search space by allowing the filter coefficients to be time-varying.In the computer simulation, the recursive digital filters are implemented in the direct form, whose filter coefficients expressed in block floating-point format are periodically time-varying (PTV). The cost function is defined by some measure of the difference between the impulse response of the synthesized filter and the desired impulse response (the Green function). The proposed scheme can be applied to the design of time-invariant and time-varying IIR filters. In synthesizing time-invariant filters, it has been demonstrated that the proposed scheme is superior to the traditional time-invariant search method under certain conditions.