Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Noise-Based Authentication: Is It Secure?

Sarah A. Flanery, Christiana Chamon

TL;DR

The paper investigates the security of noise-based authentication in a decentralized identity setting. It introduces the Chamon Authentication System (CAS), a three-layer biometric scheme using fingerprint, face, and eye-tracking noise to produce individual-specific fingerprints within a blockchain DID framework. Empirical demonstrations show that RGB-frame noise features across modalities are unique to individuals and contain tail-information that aids discrimination, but they acknowledge replication risks and propose thermal-noise alternatives. The work highlights open questions about robustness and stability under varying conditions and discusses the potential for intrinsic thermal noise to offer stronger, unspoofable authentication in future decentralized systems.

Abstract

This paper introduces a three-point biometric authentication system for a blockchain-based decentralized identity network. We use existing biometric authentication systems to demonstrate the unique noise fingerprints that belong to each individual human and the respective information leak from the biological characteristics. We then propose the concept of using unique thermal noise amplitudes generated by each user and explore the open questions regarding the robustness of unconditionally secure authentication.

Noise-Based Authentication: Is It Secure?

TL;DR

The paper investigates the security of noise-based authentication in a decentralized identity setting. It introduces the Chamon Authentication System (CAS), a three-layer biometric scheme using fingerprint, face, and eye-tracking noise to produce individual-specific fingerprints within a blockchain DID framework. Empirical demonstrations show that RGB-frame noise features across modalities are unique to individuals and contain tail-information that aids discrimination, but they acknowledge replication risks and propose thermal-noise alternatives. The work highlights open questions about robustness and stability under varying conditions and discusses the potential for intrinsic thermal noise to offer stronger, unspoofable authentication in future decentralized systems.

Abstract

This paper introduces a three-point biometric authentication system for a blockchain-based decentralized identity network. We use existing biometric authentication systems to demonstrate the unique noise fingerprints that belong to each individual human and the respective information leak from the biological characteristics. We then propose the concept of using unique thermal noise amplitudes generated by each user and explore the open questions regarding the robustness of unconditionally secure authentication.
Paper Structure (11 sections, 10 figures)

This paper contains 11 sections, 10 figures.

Table of Contents

  1. Introduction
  2. Motivation: Decentralized Identity
  3. The Importance of User Authentication
  4. Noise Fingerprinting
  5. Biometric Authentication
  6. Methodology
  7. Summary of Results
  8. Fingerprint Demonstration
  9. Face ID Demonstration
  10. Eye-Tracker Demonstration
  11. Unsolved Problem: Security of Noise-Based Authentication

Figures (10)

  • Figure 1: Example DID document for user “Kamalesh Mohanasundar” consisting of the domain of the ledger (“did:ethr”), specifically-assigned DID (“did”), and respective public key (“publicKeyHex”).
  • Figure 2: An example of (a) a user’s fingers (facing up) against a white background and (b) the resulting image after applying a mask to the skin area.
  • Figure 3: An example of (a) a photograph of a user’s face, (b) division of a photograph into frames, and (c) a single frame being assigned an RGB value.
  • Figure 4: An example of the Pupil Core eye-tracker in use. The eye-tracker rests over the user’s eyes (a) and tracks the changes in the user’s eye movement from the center of the pupil (b).
  • Figure 5: Scatter plot of the RGB frame values from the fingerprint in Figure \ref{['fingerprint_example']}, with respect to the individual frames. These values are unique to the user.
  • ...and 5 more figures