首页> 外文期刊>Pattern recognition and image analysis: advances in mathematical theory and applications in the USSR >A New Distributed Arithmetic Algorithm and Its Hardware Architecture for the Discrete Hartley Transform
【24h】

A New Distributed Arithmetic Algorithm and Its Hardware Architecture for the Discrete Hartley Transform

机译:离散Hartley变换的新型分布式算法及其硬件架构。

获取原文
获取原文并翻译 | 示例
       

摘要

This paper presents a new distributed arithmetic (DA) algorithm and its associated hardware architecture for the one-dimensional (1D) discrete Hartley transform (DHT) of any length. Using the similar idea to the Chirp-Z transform, the author derives a new algorithm that can formulate the 1D DHT of any length into cyclic convolution. By exploiting the good feature of the cyclic convolution, we can realize the 1D DHT of any length in an array architecture based on the popular DA-based approach, utilize identical ROM modules, and eliminate the accumulation loop in the processing elements (PEs). The use of identical ROM modules can simplify the ROM design process. The elimination of the accumulation loop of the PEs in conventional DA-based architecture can further increase the processing speed by applying the pipeline technique. Moreover, the presented design possesses the good features of the DA-based architectures, including low hardware cost, high computing speed, low input/output (I/O) cost, and high flexibility in transform length.
机译:本文提出了一种新的分布式算术(DA)算法及其相关的硬件体系结构,用于任意长度的一维(1D)离散Hartley变换(DHT)。利用与Chirp-Z变换类似的思想,作者得出了一种新算法,可以将任意长度的1D DHT公式化为循环卷积。通过利用循环卷积的良好特性,我们可以基于流行的基于DA的方法在阵列架构中实现任何长度的一维DHT,利用相同的ROM模块,并消除处理元件(PE)中的累积循环。使用相同的ROM模块可以简化ROM设计过程。通过应用流水线技术,消除了传统基于DA的体系结构中PE的累积环路,可以进一步提高处理速度。此外,提出的设计具有基于DA的体系结构的良好功能,包括低硬件成本,高计算速度,低输入/输出(I / O)成本以及变换长度的高度灵活性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号