一种基于TMS320LF2407A的有效FFT实现

An Effective Implementation of FFT Based on TMS320LF2407A

针对在TMS320LF2407A上直接实现快速傅利叶变换中遇到的处理速度和程序量之间的矛盾,提出一种重复算法。该算法将一维长点数快速傅立叶变换映射为两维小点数子傅立叶变换分别实现,它在保持较快处理速度的前提下减少了程序量。通过在TMS320LF2407A上对直接算法和重复算法的比较实验,发现当点数大于256时,重复算法的程序量约为直接算法的三分之一,时间上也小于直接算法,从而说明该方法的有效性和一定的实用价值。

A new iterative FFT algorithm is proposed to solve the contradiction between the program code size and the processing speed encountered in applications based on the dedicated TMSS20LF2407A. The new iterative FFT algorithm can decrease the program code size , which achieved by disassembling a long FFT to two short FFT. Compared the new iterative FFT algorithm with the direct FFT algorithm, when the number of FFT is more than 256, the program code size of the new algorithm decreases to one third of the direct algorithm programm code size and spends less time in programming. There is no deny that the new iterative FFT algorithm is effective and available.