期刊文献+

多路并行流水线型基2^(2)FFT算法实现

Implementation of Multi-path Parallel-Pipelined Radix 2^(2)FFT Algorithm
下载PDF
导出
摘要 FFT算法作为OFDM系统的核心算子占用其系统处理的大多数时间,为提高OFDM系统数据传输速度,提出了一种改进的多路并行流水线型基2^(2)FFT实现架构。在实现过程中着重对旋转因子的存储进行片上缓存优化,减少了乘法运算次数从而减小整体运算复杂度;设计的数据整合模块用于控制时序,从而保证P路并行流水型架构正确实现,数据运算吞吐率成P倍提高。RTL仿真结果表明,与同类架构相比,提出的架构在硬件开销适中的同时使得性能分别提升了1.27%、2.04%、50.88%,并且具有FFT点数可扩展的特点,可满足随着通信标准的不断提高,FFT点数逐渐增大的实际应用需求。 As the core operator of the orthogonal frequency division multiplexing(OFDM)system,the fast Fourier transform(FFT)algorithm occupies most of the system processing time.In order to improve the data transmission speed of the OFDM system,a modified multi-path parallel-pipelined radix 2^(2)FFT implementation architecture is proposed.During implementation,the on-chip cache optimization is focused on the storage of the twiddle factor,which reduces the number of multiplications and the overall computational complexity.The designed data integration module is used to control the timing,so as to ensure the correct implementation of the P-channel parallel pipeline architecture,with the throughput rate of data operation increased by P times.The RTL simulation results show that,compared with the similar architectures in three dimensions,the proposed one has a performance advantage by 1.27%,2.04%and 50.88%respectively with the moderate hardware overheads;with the great scalability of the FFT points,the proposed architecture can meet the practical application requirements for gradually increasing FFT points,with the upgrade of the communication standards.
作者 杨苗苗 郭锋 张永亮 YANG Miaomiao;GUO Feng;ZHANG Yongliang(Ordnance NCO Academy,Army Engineering University of PLA,Wuhan 430070,China)
出处 《陆军工程大学学报》 2023年第2期54-59,共6页 Journal of Army Engineering University of PLA
基金 军内科研项目(KYWHZBWX2158)。
关键词 正交频分复用 快速傅里叶变换 并行流水线 多路径延迟交叉 基2^(2) orthogonal frequency division multiplexing(OFDM) fast Fourier transform(FFT) parallel-pipelined multi-path delay commutator(MDC) radix 2^(2)
  • 相关文献

参考文献1

二级参考文献6

共引文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部