AudioStrike: Acoustic Identification of Keystrokes to Enhance End-to-End Session Integrity Público

Abstract

The lateral movement strategy is one of the most pervasive attack techniques in a modern hacker's arsenal. Generally, a point of entry is established through a phishing or social engineering attack to gain access to a target's broader network from where more confidential and valuable information is obtained. Time and time again this method of exploitation has beaten the most complex systems with state-of-the-art intrusion detection software and security infrastructure due primarily to human error. To effectively defend against lateral movement attacks, we propose Audiostrike, a continuous and frictionless keystroke authentication architecture that utilizes the natural acoustic emanations of a user's keyboard. We specifically show a proof of concept of this system on a single typist that achieves a 0.87 ROCAUC score of classifying keystrokes on three regions of the keyboard and can identify a potential attack within 5 keystrokes with high probability.

Table of Contents

1 Introduction 1

2 Background 7

2.1 Side Channel Attacks . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Security by Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.3 Threat Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.3.1 Local Compromise . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3.2 Root Compromise . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3.3 Physical Compromise . . . . . . . . . . . . . . . . . . . . . . . 10

3 Materials and Methods 12

3.1 AudioStrike System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.2 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.2.1 User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.2.2 Data Collector . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.2.3 Back-end . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.3 IRB Study Design For Crowd Sourcing . . . . . . . . . . . . . . . . . 24

3.4 Model Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.4.1 Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

4 Results 30

4.1 Data Exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

4.2 Model Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

5 Related Works 36

5.1 Security Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 36

5.2 Authentication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

5.3 Convolutional Neural Networks . . . . . . . . . . . . . . . . . . . . . 38

5.4 Audio Signal Processing . . . . . . . . . . . . . . . . . . . . . . . . . 40

5.5 Crowd Sourcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

6 Discussion 44

6.1 System Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

6.2 Ethical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 46

6.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

6.4 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

6.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Appendix A Full Keystroke Distribution 51

Bibliography 55

About this Honors Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Emory College
Department	Computer Science
Degree	B.S.
Submission	Honors Thesis
Language	English
Research Field	Computer Science
Palabra Clave	Audio Signal Processing Deep Learning Authentication Cybersecurity
Committee Chair / Thesis Advisor	Ymir Vigfusson, Emory University
Committee Members	Emily Wall, Emory University Kristin Williams, Emory University

Última modificación

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	AudioStrike: Acoustic Identification of Keystrokes to Enhance End-to-End Session Integrity ()	2023-04-06 17:11:50 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions