Agarwal-Cooley短卷积嵌套算法(ACCNA)

AGARWAL-COOLEY SHORT CONVOLUTIONS NEST ALGORITHM

离散富里叶变换(DFT)和卷积计算在图象、数字信号处理中起着重要的作用,因此对快速算法的研究早就引起人们足够的重视。自从1965年Cooley、Tukey提出基-2快速富里叶变换(FFT)算法以来,各种新算法、改进算法不断涌现,其中Winograd在1976年提出的短DFT嵌套算法(WFTA)是一种计算DFT的有效方法。1977年后,H.Silverman、J.H.McClellan、L.R.

In this paper, first we discuss Agarwal-Cooley short convolutions nest algorithm(ACCNA), and present several forms of ACCNA. Then we derive the most efficient16-point convolution algorithm; moreover, the test serial compotation of ACCNA inTurbo Pascal Language under tbe progressive Turbo-87 System on computer (lBM-PC) is given. This algorithm shown to be efficient: the time required is 65-75%less than the conventional Radix-2 FFT convolution algorithm.