Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication

Minhye Seo, Jung Yeon Hwang, Dong Hoon Lee, Soohyung Kim, Seung Hyun Kim, Jong Hwan Park

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Fuzzy authentication uses non-deterministic or noisy data, like biometrics, as an authentication factor. Although the data is extracted from the same individual or source, it can be different for each measurement. As a result, one of the main issues in fuzzy authentication is the effective processing of the fuzziness, while guaranteeing the privacy of the fuzzy data. Biometric data is a typical user-generated fuzzy data and the fuzzy extractor is one of the most promising primitives for biometric authentication these days. In 2016, Canetti et al. proposed the reusable fuzzy extractor, in which multiple keys can be generated with the same biometric data. It can also handle some outliers which occur unexpectedly (owing to an external interference when acquiring the fuzzy data, for example, the presence of dust on a fingerprint image). However, the size of the user's helper data in the reusable fuzzy extractor is quite large. This makes the network bandwidth usage required in the online authentication phase (or the storage required on the user side) considerable, which inconveniences the user. In this paper, we present a new primitive for fuzzy authentication, called a fuzzy vector signature (FVS) scheme, which significantly alleviates the burden on the user side. This means that the network bandwidth usage (or the amount of storage required on the user side) is significantly reduced. The proposed FVS scheme is reusable and robust to outliers as well. Finally, we provide a privacy-preserving fuzzy authentication protocol based on the FVS scheme.

Original languageEnglish
Article number8723318
Pages (from-to)69892-69906
Number of pages15
JournalIEEE Access
Volume7
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Authentication
Biometrics
Bandwidth
Dust
Processing

Keywords

  • Biometric authentication
  • fuzzy vector signature
  • outlier
  • privacy
  • reusability

ASJC Scopus subject areas

  • Computer Science(all)
  • Materials Science(all)
  • Engineering(all)

Cite this

Seo, M., Hwang, J. Y., Lee, D. H., Kim, S., Kim, S. H., & Park, J. H. (2019). Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication. IEEE Access, 7, 69892-69906. [8723318]. https://doi.org/10.1109/ACCESS.2019.2919351

Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication. / Seo, Minhye; Hwang, Jung Yeon; Lee, Dong Hoon; Kim, Soohyung; Kim, Seung Hyun; Park, Jong Hwan.

In: IEEE Access, Vol. 7, 8723318, 01.01.2019, p. 69892-69906.

Research output: Contribution to journalArticle

Seo, Minhye ; Hwang, Jung Yeon ; Lee, Dong Hoon ; Kim, Soohyung ; Kim, Seung Hyun ; Park, Jong Hwan. / Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication. In: IEEE Access. 2019 ; Vol. 7. pp. 69892-69906.
@article{511339fc9ce244688d977666164ad856,
title = "Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication",
abstract = "Fuzzy authentication uses non-deterministic or noisy data, like biometrics, as an authentication factor. Although the data is extracted from the same individual or source, it can be different for each measurement. As a result, one of the main issues in fuzzy authentication is the effective processing of the fuzziness, while guaranteeing the privacy of the fuzzy data. Biometric data is a typical user-generated fuzzy data and the fuzzy extractor is one of the most promising primitives for biometric authentication these days. In 2016, Canetti et al. proposed the reusable fuzzy extractor, in which multiple keys can be generated with the same biometric data. It can also handle some outliers which occur unexpectedly (owing to an external interference when acquiring the fuzzy data, for example, the presence of dust on a fingerprint image). However, the size of the user's helper data in the reusable fuzzy extractor is quite large. This makes the network bandwidth usage required in the online authentication phase (or the storage required on the user side) considerable, which inconveniences the user. In this paper, we present a new primitive for fuzzy authentication, called a fuzzy vector signature (FVS) scheme, which significantly alleviates the burden on the user side. This means that the network bandwidth usage (or the amount of storage required on the user side) is significantly reduced. The proposed FVS scheme is reusable and robust to outliers as well. Finally, we provide a privacy-preserving fuzzy authentication protocol based on the FVS scheme.",
keywords = "Biometric authentication, fuzzy vector signature, outlier, privacy, reusability",
author = "Minhye Seo and Hwang, {Jung Yeon} and Lee, {Dong Hoon} and Soohyung Kim and Kim, {Seung Hyun} and Park, {Jong Hwan}",
year = "2019",
month = "1",
day = "1",
doi = "10.1109/ACCESS.2019.2919351",
language = "English",
volume = "7",
pages = "69892--69906",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Fuzzy Vector Signature and Its Application to Privacy-Preserving Authentication

AU - Seo, Minhye

AU - Hwang, Jung Yeon

AU - Lee, Dong Hoon

AU - Kim, Soohyung

AU - Kim, Seung Hyun

AU - Park, Jong Hwan

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Fuzzy authentication uses non-deterministic or noisy data, like biometrics, as an authentication factor. Although the data is extracted from the same individual or source, it can be different for each measurement. As a result, one of the main issues in fuzzy authentication is the effective processing of the fuzziness, while guaranteeing the privacy of the fuzzy data. Biometric data is a typical user-generated fuzzy data and the fuzzy extractor is one of the most promising primitives for biometric authentication these days. In 2016, Canetti et al. proposed the reusable fuzzy extractor, in which multiple keys can be generated with the same biometric data. It can also handle some outliers which occur unexpectedly (owing to an external interference when acquiring the fuzzy data, for example, the presence of dust on a fingerprint image). However, the size of the user's helper data in the reusable fuzzy extractor is quite large. This makes the network bandwidth usage required in the online authentication phase (or the storage required on the user side) considerable, which inconveniences the user. In this paper, we present a new primitive for fuzzy authentication, called a fuzzy vector signature (FVS) scheme, which significantly alleviates the burden on the user side. This means that the network bandwidth usage (or the amount of storage required on the user side) is significantly reduced. The proposed FVS scheme is reusable and robust to outliers as well. Finally, we provide a privacy-preserving fuzzy authentication protocol based on the FVS scheme.

AB - Fuzzy authentication uses non-deterministic or noisy data, like biometrics, as an authentication factor. Although the data is extracted from the same individual or source, it can be different for each measurement. As a result, one of the main issues in fuzzy authentication is the effective processing of the fuzziness, while guaranteeing the privacy of the fuzzy data. Biometric data is a typical user-generated fuzzy data and the fuzzy extractor is one of the most promising primitives for biometric authentication these days. In 2016, Canetti et al. proposed the reusable fuzzy extractor, in which multiple keys can be generated with the same biometric data. It can also handle some outliers which occur unexpectedly (owing to an external interference when acquiring the fuzzy data, for example, the presence of dust on a fingerprint image). However, the size of the user's helper data in the reusable fuzzy extractor is quite large. This makes the network bandwidth usage required in the online authentication phase (or the storage required on the user side) considerable, which inconveniences the user. In this paper, we present a new primitive for fuzzy authentication, called a fuzzy vector signature (FVS) scheme, which significantly alleviates the burden on the user side. This means that the network bandwidth usage (or the amount of storage required on the user side) is significantly reduced. The proposed FVS scheme is reusable and robust to outliers as well. Finally, we provide a privacy-preserving fuzzy authentication protocol based on the FVS scheme.

KW - Biometric authentication

KW - fuzzy vector signature

KW - outlier

KW - privacy

KW - reusability

UR - http://www.scopus.com/inward/record.url?scp=85067255815&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067255815&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2019.2919351

DO - 10.1109/ACCESS.2019.2919351

M3 - Article

AN - SCOPUS:85067255815

VL - 7

SP - 69892

EP - 69906

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

M1 - 8723318

ER -