Human-to-Device (H2D): A Novel Anti-Tampering Mechanism for DOD Applications Driven by Cardiovascular Biometric and Obfuscation (Research Area 5.35. Hardware Assurance)

Abstract

Major Goals: The objectives for this entire research project (noted by year in brackets) were as follows: 1) ECG-based Key Generation and PPG-based User Authentication and Recognition (Year 1): While ECG-based identification had met with some success in the literature, its investigation with respect to stable, binary key generation (as required by H2D) was limited. ECG signals contain many types of noise such as baseline wander, powerline interference, electromyographic (EMG) noise, and electrode motion artifact noise as well as time-varying anomalies such as arrythmias. PPG-based biometric systems typically rely on fiducial features; that is, extraction of PPG landmarks based on time and amplitude. However, such approaches have a lower tolerance to noise, measurement conditions, and extraction algorithms. Our first objective was to quantize ECG signals into long, reliable, and high-entropy keys and to explore the accuracy of non-fiducial PPG features. This was accomplished by investigating feature extraction methods, dynamic models of ECGs/PPGs, their sources of noise, and optimized encoding algorithms. 2) Security Analysis of ECG-based Biometric Systems (Year 1) and Countermeasures (Year 3): The 7 criteria of the ideal biometric are as follows: (i) universality - possessed by all humans; (ii) distinctiveness - discriminative between individuals in the population; (iii) invariance- stable over time; (iv) collectability - quantifiably measurable; (v) performance - pertains to the availability of resources as well as achievable recognition accuracy and speed; (vi) acceptability - willingness of population to submit the attribute; and (vii) circumvention - reflects how easily a system can be fooled by a falsified biometric. While the first objective would show that how ECG could balance (i-iii) and years of ECG use in the medical field already support (iv-vi), resistance to circumvention remains uncertain.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 29, 2017
Accession Number
AD1209872

Entities

People

  • Domenic Forte

Organizations

  • University of Florida

Tags

Communities of Interest

  • Cyber
  • Energy and Power Technologies
  • Ground and Sea Platforms

DTIC Thesaurus Topics

  • Accuracy
  • Authentication
  • Circuit Boards
  • Computer Access Control
  • Databases
  • Dimensionality Reduction
  • Electrocardiography
  • Engineering
  • Feature Extraction
  • Health Services
  • Identification
  • Information Science
  • Internet Of Things
  • Presentation Attacks
  • Reliability
  • Supervised Machine Learning
  • Wearable Technology

Readers

  • Distributed Systems and Data Platform Development
  • Speech Processing/Speech Recognition.
  • Systems Analysis and Design

Technology Areas

  • AI & ML