biometric and physical identifiers with correlated noise
play

Biometric and Physical Identifiers with Correlated Noise for - PowerPoint PPT Presentation

Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication Onur G unl u Information Theory and Applications Chair, TU Berlin Joint work with Rafael F. Schaefer (TU Berlin) and H. Vincent Poor


  1. Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication Onur G¨ unl¨ u Information Theory and Applications Chair, TU Berlin Joint work with Rafael F. Schaefer (TU Berlin) and H. Vincent Poor (Princeton) ISIT 2020 G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 1

  2. <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> <latexit sha1_base64="jNlgyzleZRsxQr/wv9Grh+0L8=">ACAXicbVC7SgNBFJ31GeNr1UpsBoNgFXZFMHYBGxslgmsCyRJmJzfJkNkHM3fFsAQbf8XGQsXWv7Dzb5wkW2jigYHDOfdw54gkUKj43xbC4tLyurhbXi+sbm1ra9s3un41Rx8HgsY9UImAYpIvBQoIRGoCFgYR6MLgY+/V7UFrE0S0OE/BD1otEV3CGRmrb+y2EB8yuY6GH9AqYThWEKEete2SU3YmoPEzUmJ5Ki17a9WJ+bpOM0l07rpOgn6GVMouIRsZVqSBgfsB40DY1YCNrPJieM6JFROrQbK/MipBP1dyJjodbDMDCTIcO+nvXG4n9eM8Vuxc9ElKQIEZ8u6qaSYkzHfdCOUMBRDg1hXAnzV8r7TDGOprWiKcGdPXmeCfl87Jzc1qVvI2CuSAHJj4pIzUiWXpEY8wskjeSav5M16sl6sd+tjOrpg5Zk98gfW5w+Ld5ey</latexit> Motivation for Biometrics Noisy Measurements Source Enrollment Secret Key Helper Data Authentication Secret Key Noisy Measurements G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 1

  3. Motivation for Physical Unclonable Functions (I) Physical Unclonable Function (PUF) Q Q • Logically stable states of a static random access memory (SRAM) cell are ( Q, Q ) = (1 , 0) and (0 , 1) . • The power-up state of an SRAM cell converges to one of the states . • Uniformly random convergence due to random and uncontrollable physical mismatch of the inverters. G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 2

  4. Motivation for Physical Unclonable Functions (II) NVM= Non-Volatile Memory Secret Key Source (e.g., NVM, PUF) Encryption Decryption Key Key Decrypted Message Ciphertext Encryption Decryption Message (Plaintext) (Plaintext) G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 3

  5. Motivation for Physical Unclonable Functions (III) • Wiretap channel (WTC) communication with PUF Outputs used as the Local Source of Randomness at the WTC Encoder ALICE BOB X n Y n M channel M ENCODER DECODER Z n channel EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 4

  6. Problem Definition (I) • Suppose a secret key S is agreed securely using noisy measurements X n and Y n of the PUF outputs X n and a public communication � link , which � Allows multiple measurements via ˜ X and Y vectors; � Accounts for the effects of correlated noise in the measurements � e.g., encoder and decoder measurements are made within a coherence time; � Allows decoder measurement channels to be controllable by a cost-constrained action sequence A n ; � Hides the secret key from an eavesdropper that observes a sequence Z n correlated with ( � X n , Y n , X n ) . G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 5

  7. Problem Definition (II) ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) W ( W, S ) ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 6

  8. Problem Definition (III) - Enrollment ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) W ( W, S ) ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 6

  9. Problem Definition (III) - Enrollment ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) ( W , S ) W ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 6

  10. Problem Definition (IV) - Authentication ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) ( W, S ) W ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 7

  11. Problem Definition (IV) - Authentication ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) W ( W, S ) ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P YZ | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 7

  12. Problem Definition (IV) - Authentication ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) W ( W, S ) ˆ S = g ( W , Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 7

  13. Problem Definition (IV) - Authentication ˆ S S ( a ) ( b ) A n = f a ( W ) ( a ) = f 1 ( � X n ) W ( W, S ) ˆ S = g ( W, Y n ) ( b ) = f 2 ( � X n , S ) W � X n A n Y n P � X | X � X n P Y Z | X � XA P X X n X n Z n W EVE G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 7

  14. Problem Definition (V) ➤ Reliability: Error probability P e = Pr[ ˆ S � = S ] should vanish ; ➤ Secrecy Leakage: S should be independent of ( W, Z n ) ; ➤ Key Rate: R s = log |S| should be maximized ; n ➤ Privacy Leakage: R ℓ ≈ I ( X n ; W, Z n ) should be minimized ; n ➤ Storage: R w ≈ log |W| should be minimized ; n ➤ Action Cost: Expected action cost C ≈ E [Γ( A n )] should be minimized . G¨ unl¨ u, Schaefer, and Poor : Biometric and Physical Identifiers with Correlated Noise for Controllable Private Authentication 8

Recommend


More recommend