摘要
快速有限脉冲响应(FIR)算法(FFA)突破了传统并行FIR滤波器复杂度随并行度线性增加的局限性,效率大幅提高。然而目前缺少对高并行FFA通用算法和实现架构的研究。该文提出了高效2^n并行FFA,并给出了其通用算法形式与实现架构;同时讨论了对于非2^n并行FFA的实现架构。通过算法分析和硬件效率评估,本文算法及其实现架构在相同的并行度和性能条件下,比传统并行算法有显著改善,且随着并行度的增加,这种优势更加明显。该算法在高并行FIR滤波器的应用中有很大优势。
Fast finite impulse response(FIR)algorithm(FFA)can break the limitation that the hardware complexity of traditional parallel algorithm linearly increases with the degree of parallelism,which can improve FIR algorithm efficiency.However,there is few research focusing on the general algorithm and implementation structure for high-parallelism FFA.In this case,we propose high-efficiency 2^n-parallel general FFA and its implementation structure.We also provide the high-parallelism FFA structure with non-2^n-parallelism.According to the algorithm analysis and hardware efficiency estimation,we know that the efficient 2^n-parallel FFA algorithm and structure proposed in this paper can achieve significant advantage over traditional methods,and this advantage is becoming more obvious with the parallelism increasing.Thus,the proposed high-efficiency 2^n-parallel FFA can be used for the implementation of high-parallelism FIR filters.
作者
胡剑浩
曾维棋
费超
陈杰男
HU Jian-hao;ZENG Wei-qi;FEI Chao;CHEN Jie-nan(National Key Laboratory of Science and Technology on Communications,University of Electronic Science and Technology of China,Chengdu 611731)
出处
《电子科技大学学报》
EI
CAS
CSCD
北大核心
2020年第2期182-186,共5页
Journal of University of Electronic Science and Technology of China
基金
国家自然科学基金(61571083)。