The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
확률론적 컴퓨팅(SC)에서는 많은 확률론적 숫자(SN)를 생성해야 합니다. 일반적으로 하나의 SN을 생성하려면 LFSR(선형 피드백 시프트 레지스터)과 비교기로 구성된 SNG(확률적 숫자 생성기)가 필요합니다. SC로 산술 함수를 계산할 때 산술 함수에 사용된 상수 값과 동일한 값을 갖는 SN을 많이 생성해야 합니다. 결과적으로 하드웨어 오버헤드가 엄청납니다. 이에, 낮은 하드웨어 오버헤드로 많은 상수 SN을 생성하기 위한 GMCS(Generating Many Constant SNs from Few SNs)라는 방법이 제안되었다. 그러나 단순히 GMCS를 사용하면 생성된 상수 SN은 서로 높은 상관관계를 갖습니다. 이는 SN의 높은 상관관계로 인해 계산에 큰 오류가 발생하므로 심각한 문제가 됩니다. 따라서 본 논문에서는 오류를 증가시키지 않고 합리적으로 낮은 하드웨어 오버헤드로 일정한 SN을 생성하는 효율적인 방법을 제안합니다. GMCS에 의해 생성된 일정한 SN의 상관관계를 줄이기 위해 RRRD(Random Bit Stream Duplicator)를 사용하는 레지스터 기반 재배열 회로를 사용합니다. RRRD는 1개의 멀티플렉서(MUX)와 XNUMX개의 XNUMX비트 FF로 구성되므로 RRRD는 하드웨어 오버헤드에 거의 영향을 미치지 않습니다. 또한 하드웨어 오버헤드를 줄이기 위해 난수 생성기를 여러 SNG와 공유하는 기술을 사용합니다. 우리는 우리가 제안한 방법이 일반적으로 오류를 증가시키지 않고 일정한 SN을 생성하기 위한 하드웨어 오버헤드를 줄이는 데 매우 유용하다는 것을 확인할 수 있는 몇 가지 실험 결과를 제공합니다.
Yudai SAKAMOTO
Ritsumeikan University
Shigeru YAMASHITA
Ritsumeikan University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
부
Yudai SAKAMOTO, Shigeru YAMASHITA, "Efficient Methods to Generate Constant SNs with Considering Trade-Off between Error and Overhead and Its Evaluation" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 2, pp. 321-328, February 2020, doi: 10.1587/transinf.2018EDP7435.
Abstract: In Stochastic Computing (SC), we need to generate many stochastic numbers (SNs). If we generate one SN conventionally, we need a Stochastic Number Generator (SNG) which consists of a linear-feedback shift register (LFSR) and a comparator. When we calculate an arithmetic function by SC, we need to generate many SNs whose values are equal to constant values used in the arithmetic function. As a consequence, the hardware overhead becomes huge. Accordingly, there has been proposed a method called GMCS (Generating Many Constant SNs from Few SNs) to generate many constant SNs with low hardware overhead. However, if we use GMCS simply, generated constant SNs are correlated highly with each other. This would be a serious problem because the high correlation of SNs make a large error in computation. Therefore, in this paper, we propose efficient methods to generate constant SNs with reasonably low hardware overhead without increasing errors. To reduce the correlations of constant SNs which are generated by GMCS, we use Register based Re-arrangement circuit using a Random bit stream duplicator (RRRD). RRRDs have few influences on the hardware overhead because an RRRD consists of three multiplexers (MUXs) and two 1-bit FFs. We also use a technique to share random number generators with several SNGs to reduce the hardware overhead. We provide some experimental results by which we can confirm that our proposed methods are in general very useful to reduce the hardware overhead for generating constant SNs without increasing errors.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDP7435/_p
부
@ARTICLE{e103-d_2_321,
author={Yudai SAKAMOTO, Shigeru YAMASHITA, },
journal={IEICE TRANSACTIONS on Information},
title={Efficient Methods to Generate Constant SNs with Considering Trade-Off between Error and Overhead and Its Evaluation},
year={2020},
volume={E103-D},
number={2},
pages={321-328},
abstract={In Stochastic Computing (SC), we need to generate many stochastic numbers (SNs). If we generate one SN conventionally, we need a Stochastic Number Generator (SNG) which consists of a linear-feedback shift register (LFSR) and a comparator. When we calculate an arithmetic function by SC, we need to generate many SNs whose values are equal to constant values used in the arithmetic function. As a consequence, the hardware overhead becomes huge. Accordingly, there has been proposed a method called GMCS (Generating Many Constant SNs from Few SNs) to generate many constant SNs with low hardware overhead. However, if we use GMCS simply, generated constant SNs are correlated highly with each other. This would be a serious problem because the high correlation of SNs make a large error in computation. Therefore, in this paper, we propose efficient methods to generate constant SNs with reasonably low hardware overhead without increasing errors. To reduce the correlations of constant SNs which are generated by GMCS, we use Register based Re-arrangement circuit using a Random bit stream duplicator (RRRD). RRRDs have few influences on the hardware overhead because an RRRD consists of three multiplexers (MUXs) and two 1-bit FFs. We also use a technique to share random number generators with several SNGs to reduce the hardware overhead. We provide some experimental results by which we can confirm that our proposed methods are in general very useful to reduce the hardware overhead for generating constant SNs without increasing errors.},
keywords={},
doi={10.1587/transinf.2018EDP7435},
ISSN={1745-1361},
month={February},}
부
TY - JOUR
TI - Efficient Methods to Generate Constant SNs with Considering Trade-Off between Error and Overhead and Its Evaluation
T2 - IEICE TRANSACTIONS on Information
SP - 321
EP - 328
AU - Yudai SAKAMOTO
AU - Shigeru YAMASHITA
PY - 2020
DO - 10.1587/transinf.2018EDP7435
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2020
AB - In Stochastic Computing (SC), we need to generate many stochastic numbers (SNs). If we generate one SN conventionally, we need a Stochastic Number Generator (SNG) which consists of a linear-feedback shift register (LFSR) and a comparator. When we calculate an arithmetic function by SC, we need to generate many SNs whose values are equal to constant values used in the arithmetic function. As a consequence, the hardware overhead becomes huge. Accordingly, there has been proposed a method called GMCS (Generating Many Constant SNs from Few SNs) to generate many constant SNs with low hardware overhead. However, if we use GMCS simply, generated constant SNs are correlated highly with each other. This would be a serious problem because the high correlation of SNs make a large error in computation. Therefore, in this paper, we propose efficient methods to generate constant SNs with reasonably low hardware overhead without increasing errors. To reduce the correlations of constant SNs which are generated by GMCS, we use Register based Re-arrangement circuit using a Random bit stream duplicator (RRRD). RRRDs have few influences on the hardware overhead because an RRRD consists of three multiplexers (MUXs) and two 1-bit FFs. We also use a technique to share random number generators with several SNGs to reduce the hardware overhead. We provide some experimental results by which we can confirm that our proposed methods are in general very useful to reduce the hardware overhead for generating constant SNs without increasing errors.
ER -