FITness Assessment - Hardware Algorithm Safety Validation

Andreas Strasser, Philipp Stelzer, Christian Steger, Norbert Druml

Research output: Contribution to conference › Paper › Research › peer-review

Abstract

Error correction codes (ECC) are important safety methods for digital data to gain control of single event upsets (SEU) in integrated digital circuits. SEU are responsible for single bit flips inside a digital circuit caused by ionizing radiation. This effect does not affect the physical structure of the components but the correctness of data inside flip flops. Consequently, data gets corrupted and the correct program flow gets disturbed. This effect needs to be considered especially for safety-critical systems. In the novel ISO 26262 2nd Edition, the automotive domain suggests controlling SEU effects by algorithms that correct single bit errors and detect double bit errors (SEC-DED). This raises the question what kind of impact double bit error correction (DEC) will have on the overall safety level for LiDAR systems. In this paper we determine the difference between two ECC algorithms from a safety point of view: Hamming's code (SEC-DED) and BCH-code (DEC). For this purpose we developed a novel method for algorithm safety validation and applied it to both algorithms.
Original languageEnglish
Number of pages6
Publication statusPublished - 24 Mar 2019
EventThe Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications - Valencia, Spain
Duration: 24 Mar 2019 → 28 Mar 2019

Conference

ConferenceThe Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications
Abbreviated titlePESARO
CountrySpain
CityValencia
Period24/03/19 → 28/03/19

Fingerprint

Hardware
Error correction
Codes (standards)
Digital integrated circuits
Flip flop circuits
Gain control
Ionizing radiation
Digital circuits

Keywords

    Cite this

    Strasser, A., Stelzer, P., Steger, C., & Druml, N. (2019). FITness Assessment - Hardware Algorithm Safety Validation. Paper presented at The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, Valencia, Spain.

    FITness Assessment - Hardware Algorithm Safety Validation. / Strasser, Andreas; Stelzer, Philipp; Steger, Christian; Druml, Norbert.

    2019. Paper presented at The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, Valencia, Spain.

    Research output: Contribution to conference › Paper › Research › peer-review

    Strasser, A, Stelzer, P, Steger, C & Druml, N 2019, 'FITness Assessment - Hardware Algorithm Safety Validation' Paper presented at The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, Valencia, Spain, 24/03/19 - 28/03/19, .
    Strasser A, Stelzer P, Steger C, Druml N. FITness Assessment - Hardware Algorithm Safety Validation. 2019. Paper presented at The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, Valencia, Spain.
    Strasser, Andreas ; Stelzer, Philipp ; Steger, Christian ; Druml, Norbert. / FITness Assessment - Hardware Algorithm Safety Validation. Paper presented at The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, Valencia, Spain.6 p.
    @conference{49ced53d03d645b891be85ce8a191104,
    title = "FITness Assessment - Hardware Algorithm Safety Validation",
    abstract = "Error correction codes (ECC) are important safety methods for digital data to gain control of single event upsets (SEU) in integrated digital circuits. SEU are responsible for single bit flips inside a digital circuit caused by ionizing radiation. This effect does not affect the physical structure of the components but the correctness of data inside flip flops. Consequently, data gets corrupted and the correct program flow gets disturbed. This effect needs to be considered especially for safety-critical systems. In the novel ISO 26262 2nd Edition, the automotive domain suggests controlling SEU effects by algorithms that correct single bit errors and detect double bit errors (SEC-DED). This raises the question what kind of impact double bit error correction (DEC) will have on the overall safety level for LiDAR systems. In this paper we determine the difference between two ECC algorithms from a safety point of view: Hamming's code (SEC-DED) and BCH-code (DEC). For this purpose we developed a novel method for algorithm safety validation and applied it to both algorithms.",
    keywords = "FIT Estimation, Error Correction Codes, LiDAR",
    author = "Andreas Strasser and Philipp Stelzer and Christian Steger and Norbert Druml",
    year = "2019",
    month = "3",
    day = "24",
    language = "English",
    note = "The Ninth International Conference on Performance, Safety and Robustness in Complex Systems and Applications, PESARO ; Conference date: 24-03-2019 Through 28-03-2019",

    }

    TY - CONF

    T1 - FITness Assessment - Hardware Algorithm Safety Validation

    AU - Strasser, Andreas

    AU - Stelzer, Philipp

    AU - Steger, Christian

    AU - Druml, Norbert

    PY - 2019/3/24

    Y1 - 2019/3/24

    N2 - Error correction codes (ECC) are important safety methods for digital data to gain control of single event upsets (SEU) in integrated digital circuits. SEU are responsible for single bit flips inside a digital circuit caused by ionizing radiation. This effect does not affect the physical structure of the components but the correctness of data inside flip flops. Consequently, data gets corrupted and the correct program flow gets disturbed. This effect needs to be considered especially for safety-critical systems. In the novel ISO 26262 2nd Edition, the automotive domain suggests controlling SEU effects by algorithms that correct single bit errors and detect double bit errors (SEC-DED). This raises the question what kind of impact double bit error correction (DEC) will have on the overall safety level for LiDAR systems. In this paper we determine the difference between two ECC algorithms from a safety point of view: Hamming's code (SEC-DED) and BCH-code (DEC). For this purpose we developed a novel method for algorithm safety validation and applied it to both algorithms.

    AB - Error correction codes (ECC) are important safety methods for digital data to gain control of single event upsets (SEU) in integrated digital circuits. SEU are responsible for single bit flips inside a digital circuit caused by ionizing radiation. This effect does not affect the physical structure of the components but the correctness of data inside flip flops. Consequently, data gets corrupted and the correct program flow gets disturbed. This effect needs to be considered especially for safety-critical systems. In the novel ISO 26262 2nd Edition, the automotive domain suggests controlling SEU effects by algorithms that correct single bit errors and detect double bit errors (SEC-DED). This raises the question what kind of impact double bit error correction (DEC) will have on the overall safety level for LiDAR systems. In this paper we determine the difference between two ECC algorithms from a safety point of view: Hamming's code (SEC-DED) and BCH-code (DEC). For this purpose we developed a novel method for algorithm safety validation and applied it to both algorithms.

    KW - FIT Estimation

    KW - Error Correction Codes

    KW - LiDAR

    M3 - Paper

    ER -

    Related content

    Projects

    Hardware/Software-Codesign

    Project: Research area

    PRYSTINE - Programmable Systems for Intelligence in Automobiles

    Project: Research project