循环卷积DFT的优化算法与仿真

The optimization and simulation of circular convolution DFT

根据余数系统中模映射法则以及数论变换,将素数N点的DFT运算转换为N-1点的循环卷积运算,建立了算法模型,给出了此算法的FIR滤波器图解,并对加法器系数进行RAG优化,最后在Model Sim仿真平台上,用Verilog语言实现该算法,并进行了仿真结果分析和工作量分析。RAG优化后减少了加法器数量,降低了路径延迟。

In this paper,according to the theorem of mold mapping rule in Residue Number System（ RNS） and number theoretic transform,the prime number N-point Discrete Fourier Transform（ DFT） is converted to N-1 point circular convolution operation. And algorithm model is established,corresponding Finite Impulse Response（ FIR） filter illustrated is given by using RAG optimized coefficients. Finally the algorithm is implemented by using Verilog in Model Sim simulation platform and the simulating result and analysis are given. It concluded that the number of adders and path delay is reduced.