ROUNDOFF ERRORS IN FIXED-POINT DIGITAL FILTERS

摘要：

When operating, the computational module of digital filters performs only multiplications and additions. In most hardware implementations, these arithmetic operations are performed in fixed-point. An immediate issue regarding the implementation of the multiplier is that some form of quantization scheme must be included in the hardware to convert the double-wordlength pruduct term back to single-wordlength. The most commonly used quantization schemes are rounding and tmncatinn. However, both schemes will lead to a DC bias in the filter output. In this paper, we consider a zero-bias quantizatinn scheme and study the probability distribution and spectral characteristics of the resultingroundoff errors.

关键词： Fized-point arithmetic Roundofferrors IIR filters

作者: KIM YONG TAN YEE HONG LEUNG

作者单位: Depertmeat of Electrical and Computer Engineering, Curtin University of Technology, Bentley WA, Australia

会议类型: 国际会议

会议名称: The Second International Conference on Information & Systems Sciences(ICISS2008)(第二届信息与系统科学国际会议)

会议地点: 大连

会议语种:英文

页码: 611-619

在线出版日期: 2008-12-18（万方平台首次上网日期，不代表论文的发表时间）