Stochastic number duplicators based on bit re-arrangement using randomized bit streams

    Research output: Contribution to journalArticle

    3 Citations (Scopus)

    Abstract

    Recently, stochastic computing based on stochastic numbers attracts attention as an effective computation method, which realizes arithmetic operations by simple logic circuits with a tolerance of bit errors. When we input two or more identical values to a stochastic circuit, we require to duplicate a stochastic number. However, if bit streams of duplicated stochastic numbers are dependent on each other, their arithmetic operation results can be inaccurate. In this paper, we propose two stochastic number duplicators, called FSR and RRR. The stochastic numbers duplicated by the FSR and RRR duplicators have the equivalent values but have independent bit streams, effectively utilizing bit re-arrangement using randomized bit streams. Experimental evaluation results demonstrate that the RRR duplicator, in particular, obtains more accurate results even if a circuit has re-convergence paths, reducing the mean square errors by 20%-89% compared to a conventional stochastic number duplicator.

    Original languageEnglish
    Pages (from-to)1002-1013
    Number of pages12
    JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
    VolumeE101A
    Issue number7
    DOIs
    Publication statusPublished - 2018 Jul 1

    Fingerprint

    Rearrangement
    Networks (circuits)
    Logic circuits
    Mean square error
    Inaccurate
    Experimental Evaluation
    Tolerance
    Logic
    Path
    Dependent
    Computing
    Demonstrate

    Keywords

    • Bit rearrangement
    • Duplicator
    • Re-convergence path
    • Stochastic computing
    • Stochastic number

    ASJC Scopus subject areas

    • Signal Processing
    • Computer Graphics and Computer-Aided Design
    • Electrical and Electronic Engineering
    • Applied Mathematics

    Cite this

    @article{ec7a71fb19cd4df4bc70217c106f8a84,
    title = "Stochastic number duplicators based on bit re-arrangement using randomized bit streams",
    abstract = "Recently, stochastic computing based on stochastic numbers attracts attention as an effective computation method, which realizes arithmetic operations by simple logic circuits with a tolerance of bit errors. When we input two or more identical values to a stochastic circuit, we require to duplicate a stochastic number. However, if bit streams of duplicated stochastic numbers are dependent on each other, their arithmetic operation results can be inaccurate. In this paper, we propose two stochastic number duplicators, called FSR and RRR. The stochastic numbers duplicated by the FSR and RRR duplicators have the equivalent values but have independent bit streams, effectively utilizing bit re-arrangement using randomized bit streams. Experimental evaluation results demonstrate that the RRR duplicator, in particular, obtains more accurate results even if a circuit has re-convergence paths, reducing the mean square errors by 20{\%}-89{\%} compared to a conventional stochastic number duplicator.",
    keywords = "Bit rearrangement, Duplicator, Re-convergence path, Stochastic computing, Stochastic number",
    author = "Ryota Ishikawa and Masashi Tawada and Masao Yanagisawa and Nozomu Togawa",
    year = "2018",
    month = "7",
    day = "1",
    doi = "10.1587/transfun.E101.A.1002",
    language = "English",
    volume = "E101A",
    pages = "1002--1013",
    journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",
    issn = "0916-8508",
    publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
    number = "7",

    }

    TY - JOUR

    T1 - Stochastic number duplicators based on bit re-arrangement using randomized bit streams

    AU - Ishikawa, Ryota

    AU - Tawada, Masashi

    AU - Yanagisawa, Masao

    AU - Togawa, Nozomu

    PY - 2018/7/1

    Y1 - 2018/7/1

    N2 - Recently, stochastic computing based on stochastic numbers attracts attention as an effective computation method, which realizes arithmetic operations by simple logic circuits with a tolerance of bit errors. When we input two or more identical values to a stochastic circuit, we require to duplicate a stochastic number. However, if bit streams of duplicated stochastic numbers are dependent on each other, their arithmetic operation results can be inaccurate. In this paper, we propose two stochastic number duplicators, called FSR and RRR. The stochastic numbers duplicated by the FSR and RRR duplicators have the equivalent values but have independent bit streams, effectively utilizing bit re-arrangement using randomized bit streams. Experimental evaluation results demonstrate that the RRR duplicator, in particular, obtains more accurate results even if a circuit has re-convergence paths, reducing the mean square errors by 20%-89% compared to a conventional stochastic number duplicator.

    AB - Recently, stochastic computing based on stochastic numbers attracts attention as an effective computation method, which realizes arithmetic operations by simple logic circuits with a tolerance of bit errors. When we input two or more identical values to a stochastic circuit, we require to duplicate a stochastic number. However, if bit streams of duplicated stochastic numbers are dependent on each other, their arithmetic operation results can be inaccurate. In this paper, we propose two stochastic number duplicators, called FSR and RRR. The stochastic numbers duplicated by the FSR and RRR duplicators have the equivalent values but have independent bit streams, effectively utilizing bit re-arrangement using randomized bit streams. Experimental evaluation results demonstrate that the RRR duplicator, in particular, obtains more accurate results even if a circuit has re-convergence paths, reducing the mean square errors by 20%-89% compared to a conventional stochastic number duplicator.

    KW - Bit rearrangement

    KW - Duplicator

    KW - Re-convergence path

    KW - Stochastic computing

    KW - Stochastic number

    UR - http://www.scopus.com/inward/record.url?scp=85049391236&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85049391236&partnerID=8YFLogxK

    U2 - 10.1587/transfun.E101.A.1002

    DO - 10.1587/transfun.E101.A.1002

    M3 - Article

    AN - SCOPUS:85049391236

    VL - E101A

    SP - 1002

    EP - 1013

    JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

    JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

    SN - 0916-8508

    IS - 7

    ER -