發(fā)布時間：2018-01-18 06:15

  本文關(guān)鍵詞：高能效混合浮點FFT硬件加速器架構(gòu)與VLSI實現(xiàn)研究　出處：《復(fù)旦大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 快速傅里葉變換 高能效電路 低成本電路 混合浮點

【摘要】：快速傅里葉變換(FFT)是數(shù)字信號處理中最常用的算法之一。它始終是數(shù)字信號處理領(lǐng)域的研究熱點。如今,FFT是很多新興應(yīng)用中的關(guān)鍵處理模塊,如基于正交頻分復(fù)用(OFDM)的手持移動通信系統(tǒng)和生物醫(yī)療電子信號處理平臺。這些應(yīng)用有一個顯著的共同點,那就是它們要求整個系統(tǒng)的功耗極低,以延長產(chǎn)品的使用周期。同時,它們也要求系統(tǒng)具備良好的適應(yīng)性,在面對不同信號輸入時,都能給出理想的處理結(jié)果。因此,FFT硬件加速器必需在保證一定量化信噪比(SQNR)輸出的前提下做到高能效、低成本和高靈活性的實現(xiàn)。針對上述要求,本文從算法和電路層面優(yōu)化設(shè)計實現(xiàn)FFT硬件加速器。在算法方面,本文總結(jié)了FFT硬件實現(xiàn)中常用的數(shù)據(jù)表示格式,包括定點格式、浮點格式和基于定點縮放的方法。在這些格式的基礎(chǔ)上,本文提出了動態(tài)偏置調(diào)節(jié)的混合浮點方法。該方法采用浮點格式的指數(shù)域和定點格式的小數(shù)域,并使復(fù)數(shù)的實部和虛部共享一個指數(shù)域。這樣可以在保證數(shù)據(jù)精度的前提下,減少硬件實現(xiàn)的成本和功耗。此外,動態(tài)偏置調(diào)節(jié)的方法可以根據(jù)輸入信號的不同在運算過程中動態(tài)調(diào)整數(shù)據(jù)表示范圍,從而提高整體SQNR。這種機(jī)制保證了FFT硬件加速器的靈活性和高精度輸出。因此,采用動態(tài)偏置調(diào)節(jié)的混合浮點方法的FFT硬件加速器能夠以較小數(shù)據(jù)位寬獲得較高SQNR,從而達(dá)到降低功耗和成本的目標(biāo)。在電路層面,本文實現(xiàn)的FFT硬件加速器采用單存儲器架構(gòu)以降低硬件的開銷。在數(shù)據(jù)通路的實現(xiàn)中,本文采用多種方法來降低功耗和提高SQNR。第一,本文分析并減少蝶形運算中所需的浮點歸一化操作,由原來的15個操作降低到4個操作。第二,本文分析并縮短蝶形運算中所需的數(shù)據(jù)處理位寬,在小數(shù)位寬為9時,可以使中間處理位寬節(jié)省多達(dá)6比特。第三,本文采用Trounding的數(shù)據(jù)舍去策略,盡可能地降低量化誤差而不增加過多的硬件開銷。此外,本文最后著眼于基于低電壓存儲器的FFT硬件加速器設(shè)計。首先概述存儲器故障的種類和產(chǎn)生原因。然后描述了一定電壓下存儲器故障率的分析仿真方法。之后,給出具體故障率與電壓和電路頻率之間的關(guān)系。并根據(jù)這個對應(yīng)關(guān)系分析出一定存儲器電壓下FFT硬件加速器的SQNR以及該情況下的功耗收益。本文提出的FFT硬件加速器能夠計算64-8192點的變換。當(dāng)數(shù)據(jù)位寬為3+2*9比特,存儲器電壓為0.7V,使用SMIC 65nm工藝時,FFT硬件加速器工作在400MHz,面積為0.482mm2,功耗為35.3mW。64點和8192點對應(yīng)的SQNR分別為41.6 dB和35.8 dB。
[Abstract]:Fast Fourier transform (FFT) is one of the most commonly used algorithms in digital signal processing. It has always been a research hotspot in the field of digital signal processing. Nowadays, FFT is a key processing module in many new applications. For example, the handheld mobile communication system and biomedical electronic signal processing platform based on OFDM (orthogonal Frequency Division Multiplexing). These applications have a remarkable common point, that is, they require very low power consumption of the whole system. At the same time, they also require the system to have good adaptability, in the face of different signal input, can give the ideal processing results. FFT hardware accelerator must achieve high energy efficiency, low cost and high flexibility on the premise of certain quantization signal-to-noise ratio (SNR) output. This paper optimizes the design and implementation of FFT hardware accelerator from the algorithm and circuit level. In the aspect of algorithm, this paper summarizes the data representation format commonly used in FFT hardware implementation, including fixed-point format. On the basis of these formats, a mixed floating point method for dynamic bias adjustment is proposed. This method uses the exponential domain of floating point format and the decimal domain of fixed point format. The real and virtual parts of the complex can share an exponential domain, which can reduce the cost and power consumption of the hardware implementation under the premise of ensuring the data accuracy. The method of dynamic bias adjustment can dynamically adjust the range of data representation according to the difference of input signal in the operation process. This mechanism ensures the flexibility and high precision output of FFT hardware accelerator. The hybrid floating-point accelerator with dynamic bias adjustment can obtain higher SQNRs with smaller data bit width, thus achieving the goal of reducing power consumption and cost at the circuit level. The hardware accelerator implemented in this paper uses single memory architecture to reduce hardware overhead. In the implementation of data path, this paper uses a variety of methods to reduce power consumption and improve SQNR. first. This paper analyzes and reduces the floating point normalization operation in butterfly operation, from 15 operations to 4 operations. Secondly, this paper analyzes and shortens the bit width of data processing required in butterfly operation. When the decimal width is 9:00, the intermediate processing bit width can be saved by up to 6 bits. Thirdly, this paper adopts the data reduction strategy of Trounding. Minimize quantization errors without adding too much hardware overhead. In the end, this paper focuses on the design of FFT hardware accelerator based on low voltage memory. Firstly, the types and causes of memory failure are summarized. Then, the analysis and simulation method of memory failure rate under certain voltage is described. After. The relationship between failure rate, voltage and circuit frequency is given. According to this relation, the SQNR of FFT hardware accelerator under certain memory voltage and the power gain in this case are analyzed. The FFT hardware accelerator can calculate the 64-8192 point transformation. When the data bit width is 3. Two or nine bits. The memory voltage is 0.7 V, and the SMIC hardware accelerator is working at 400MHz with an area of 0.482mm2 using SMIC 65nm process. The SQNR corresponding to the power consumption of 35.3mW.64 and 8192 is 41.6 dB and 35.8 dB, respectively.
【學(xué)位授予單位】：復(fù)旦大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TN911.72;TN47

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 顧葉華;曾曉洋;韓軍;張章;;低硬件成本的高性能Hash算法加速器VLSI設(shè)計[J];小型微型計算機(jī)系統(tǒng);2007年05期

2 張啟英;劉亞剛;張淑艷;朱娟;;基于FPGA的硬件加速器設(shè)計的研究與應(yīng)用[J];計算機(jī)光盤軟件與應(yīng)用;2013年17期

3 周凡;時龍興;楊軍;;基于統(tǒng)計分析的SoC定點硬件加速器字長設(shè)計[J];固體電子學(xué)研究與進(jìn)展;2007年02期

4 裴明敬;張歆奕;;基于Nios Ⅱ的IMDCT算法硬件加速器設(shè)計[J];五邑大學(xué)學(xué)報(自然科學(xué)版);2014年01期

5 Noam Shendar;;Java軟件解決方案是怎樣勝過硬件加速器的?[J];集成電路應(yīng)用;2005年06期

6 楊煉;彭濤;;EPS系統(tǒng)中業(yè)務(wù)流模板硬件加速器設(shè)計[J];電視技術(shù);2012年23期

7 ;愛國者推出新一代MX31移動多媒體技術(shù)平臺[J];世界電子元器件;2007年01期

8 朱燕翔;周凡;;MP3解碼的IMDCT硬件加速器方案[J];單片機(jī)與嵌入式系統(tǒng)應(yīng)用;2006年11期

9 周凡;時龍興;楊軍;張宇;高谷剛;;浮定點轉(zhuǎn)換與SoC定點加速器字長協(xié)同設(shè)計研究[J];應(yīng)用科學(xué)學(xué)報;2007年02期

10 孫宇豪;張盛兵;;基于FPGA的ROHC硬件解壓器設(shè)計[J];微電子學(xué)與計算機(jī);2013年08期

相關(guān)重要報紙文章 前1條

1 本報記者 趙艷秋;基站芯片：積極部署3G/LTE 頻推高性能方案[N];中國電子報;2009年

相關(guān)碩士學(xué)位論文 前5條

1 席毅;基于硬件加速器的DMC控制器實現(xiàn)研究[D];吉林大學(xué);2011年

2 王振;LTE終端加解密硬件加速器的研究與設(shè)計[D];西安科技大學(xué);2012年

3 薄一帆;高能效混合浮點FFT硬件加速器架構(gòu)與VLSI實現(xiàn)研究[D];復(fù)旦大學(xué);2014年

4 庹偉;醫(yī)用CT三維重建系統(tǒng)中PCIE數(shù)據(jù)傳輸接口的FPGA設(shè)計與實現(xiàn)[D];哈爾濱工業(yè)大學(xué);2008年

5 趙開蘭;靈活可配的大數(shù)運算架構(gòu)設(shè)計[D];浙江大學(xué);2014年

，

本文編號：1439788