基于Catapult C工具递归型滤波器设计和算法优化  

Research on Catapult C tool and optimization algorithm of recursive filter

作　　者：叶海雄 陶宁蓉[1] 王世明[1] 

机构地区：[1]上海海洋大学工程学院

出　　处：《电子设计工程》2017年第22期10-14,共5页Electronic Design Engineering

基　　金：国家自然科学基金项目(71501125);上海市青年教师高校新进教师培训及科研启动基金(ZZHY14033)

摘　　要：使用高层次综合方法设计超大规模集成电路是一项前瞻性的工作。其设计理念旨在保证电路性能的前提下,缩短电路开发周期,让产品更早投入到市场。然而针对高层次综合工具面对算法转换时编译能力存在局限的问题,本文以信号处理中递归型滤波器为例,针对滤波器功耗的问题,通过分析滤波算法转换以及评估高层次综合工具对滤波器的影响。实验结果表明,采用算法转换及优化的方法,能使电路面积减少58%,能耗降低25%,达到低碳化、节能环保的电路设计要求。Using High-Level Synthesis(HLS) method to design Very Large Scale Integrated(VLSI)circuit is an advanced work. The design philosophy is to shorten the developing time and accelerate the time to market under guarantying the performance. However the problem limited the compiler of HighLevel synthesis tool to algorithm transformation optimization, this paper based on the Infinite Impulse Response(IIR)recursive filter in signal processing as an example, we analyzed and evaluated the impacts on the algorithm optimization transformation to the HLS tool Catapult C. The experimental results showed the circuit 58% on area and 25% on energy consumption, meeting the requirements of low carbon and ecologic energy.

关 键 词：高层次综合 集成电路 信号处理 递归型滤波器 面积 能耗 

分 类 号：TN4[电子电信—微电子学与固体电子学]