anonymity networks
play

Anonymity Networks Laslo Hunhold Mathematisches Institut - PowerPoint PPT Presentation

Anonymity Networks Laslo Hunhold Mathematisches Institut Universitt zu Kln 27th July 2017 In the lecture Information Theory and Statistical Physics by Prof. Dr. Johannes Berg motivation motivation hide initiator of a message in a


  1. Anonymity Networks Laslo Hunhold Mathematisches Institut Universität zu Köln 27th July 2017 In the lecture ‘Information Theory and Statistical Physics’ by Prof. Dr. Johannes Berg

  2. motivation

  3. motivation ◮ hide initiator of a message in a computer network

  4. motivation ◮ hide initiator of a message in a computer network ◮ safe whistleblowing under corporate and state surveillance

  5. motivation ◮ hide initiator of a message in a computer network ◮ safe whistleblowing under corporate and state surveillance ◮ ‘deniable communication’

  6. motivation ◮ hide initiator of a message in a computer network ◮ safe whistleblowing under corporate and state surveillance ◮ ‘deniable communication’ ◮ decentralized

  7. idea

  8. idea node network participant link possible message path

  9. idea node network participant link possible message path ◮ all nodes have equal weight

  10. idea node network participant link possible message path ◮ all nodes have equal weight ◮ message unmodifiable, only receiver is known

  11. idea node network participant link possible message path ◮ all nodes have equal weight ◮ message unmodifiable, only receiver is known ◮ each node on path: biased coin flip: forward or deliver

  12. idea node network participant link possible message path ◮ all nodes have equal weight ◮ message unmodifiable, only receiver is known ◮ each node on path: biased coin flip: forward or deliver ◮ each node on path: initiator or just forwarder?

  13. idea node network participant link possible message path ◮ all nodes have equal weight ◮ message unmodifiable, only receiver is known ◮ each node on path: biased coin flip: forward or deliver ◮ each node on path: initiator or just forwarder? → message initator gets lost in the crowd

  14. model

  15. model n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ N nodes n 1 , . . . , n N with P ( n i is initiator) =: P ( X = n i ) =: p i

  16. model n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ N nodes n 1 , . . . , n N with P ( n i is initiator) =: P ( X = n i ) =: p i ◮ n i probably innocent ↔ p i ≤ 1 2

  17. model n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ N nodes n 1 , . . . , n N with P ( n i is initiator) =: P ( X = n i ) =: p i ◮ n i probably innocent ↔ p i ≤ 1 2 ◮ forwarding probability λ

  18. model n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ N nodes n 1 , . . . , n N with P ( n i is initiator) =: P ( X = n i ) =: p i ◮ n i probably innocent ↔ p i ≤ 1 2 ◮ forwarding probability λ if message received then flip biased coin P (heads) = λ if heads then forward to a uniformly chosen node else deliver to receiver end if end if

  19. degree of anonymity

  20. degree of anonymity best case X := X : ∀ i ∈ { 1 , . . . , N } : p i = 1 N

  21. degree of anonymity best case X := X : ∀ i ∈ { 1 , . . . , N } : p i = 1 N N � H := H ( X ) = − p i · ln( p i ) = ln( N − C ) i =1

  22. degree of anonymity best case X := X : ∀ i ∈ { 1 , . . . , N } : p i = 1 N N � H := H ( X ) = − p i · ln( p i ) = ln( N − C ) i =1 worst case X := X : ∀ i ∈ { 1 , . . . , N } \ { j } : p i = 0 ∧ p j = 1

  23. degree of anonymity best case X := X : ∀ i ∈ { 1 , . . . , N } : p i = 1 N N � H := H ( X ) = − p i · ln( p i ) = ln( N − C ) i =1 worst case X := X : ∀ i ∈ { 1 , . . . , N } \ { j } : p i = 0 ∧ p j = 1 N � H := H ( X ) = − p i · ln( p i ) = 1 · ln(1) = 0 i =1

  24. degree of anonymity best case X := X : ∀ i ∈ { 1 , . . . , N } : p i = 1 N N � H := H ( X ) = − p i · ln( p i ) = ln( N − C ) i =1 worst case X := X : ∀ i ∈ { 1 , . . . , N } \ { j } : p i = 0 ∧ p j = 1 N � H := H ( X ) = − p i · ln( p i ) = 1 · ln(1) = 0 i =1 d ( X ) := 1 − H − H ( X ) = H ( X ) ∈ [0 , 1] H H

  25. corruption

  26. corruption n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ 0 ≤ C < N corrupt nodes (incoming message passer known)

  27. corruption n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ 0 ≤ C < N corrupt nodes (incoming message passer known) ◮ behave normally

  28. corruption n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ 0 ≤ C < N corrupt nodes (incoming message passer known) ◮ behave normally ◮ wait for message to be passed to us

  29. corruption n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ 0 ≤ C < N corrupt nodes (incoming message passer known) ◮ behave normally ◮ wait for message to be passed to us ◮ analyze probability of passer being initiator

  30. corruption n 4 n 5 n 3 n 6 n 2 n 7 n 1 n 8 ◮ 0 ≤ C < N corrupt nodes (incoming message passer known) ◮ behave normally ◮ wait for message to be passed to us ◮ analyze probability of passer being initiator P (passer is initiator) > 1 2 → unmasked

  31. analysis events

  32. analysis events 4 n I n C 3 1 n n n m 2

  33. analysis events 4 n I n C 3 1 n n n m 2 let k > 0

  34. analysis events 4 n I n C 3 1 n n n m 2 let k > 0 H k := first corrupt node is at the k th path-position

  35. analysis events 4 n I n C 3 1 n n n m 2 let k > 0 H k := first corrupt node is at the k th path-position ∞ � H k + := H i i = k

  36. analysis events 4 n I n C 3 1 n n n m 2 let k > 0 H k := first corrupt node is at the k th path-position ∞ � H k + := H i i = k I := first corrupt node immediately postcedes the message initiator

  37. analysis events 4 n I n C 3 1 n n n m 2 let k > 0 H k := first corrupt node is at the k th path-position ∞ � H k + := H i i = k I := first corrupt node immediately postcedes the message initiator P (passer is initiator) = P ( I | H 1+ )

  38. analysis events 4 n I n C 3 1 n n n m 2 let k > 0 H k := first corrupt node is at the k th path-position ∞ � H k + := H i i = k I := first corrupt node immediately postcedes the message initiator P (passer is initiator) = P ( I | H 1+ ) note: H 1 → I , but I �→ H 1

  39. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N

  40. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N proof:

  41. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N proof: � k − 1 � λ · N − C � λ · C � P ( H k ) = · N N

  42. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N proof: � k − 1 � λ · N − C � λ · C � P ( H k ) = · N N � k � λ · N − C ∞ C · N � ⇒ P ( H k + ) = P ( H i ) = . . . = � � 1 − λ · N − C ( N − C ) · i = k N

  43. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N proof: � k − 1 � λ · N − C � λ · C � P ( H k ) = · N N � k � λ · N − C ∞ C · N � ⇒ P ( H k + ) = P ( H i ) = . . . = � � 1 − λ · N − C ( N − C ) · i = k N H 1 → I ⇒ P ( I | H 1 ) = 1

  44. analysis general probability I P ( I | H 1+ ) = N − λ ( N − C − 1) N proof: � k − 1 � λ · N − C � λ · C � P ( H k ) = · N N � k � λ · N − C ∞ C · N � ⇒ P ( H k + ) = P ( H i ) = . . . = � � 1 − λ · N − C ( N − C ) · i = k N H 1 → I ⇒ P ( I | H 1 ) = 1 1 P ( I | H 2+ ) = N − C

  45. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ )

  46. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . .

  47. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N

  48. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N = P ( I ∧ H 1+ ) P ( I | H 1+ ) CP P ( H 1+ )

  49. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N = P ( I ∧ H 1+ ) P ( I | H 1+ ) CP � I → H 1+ � P ( H 1+ ) P ( I ) = P ( H 1+ )

  50. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N = P ( I ∧ H 1+ ) P ( I | H 1+ ) CP � I → H 1+ � P ( H 1+ ) P ( I ) = P ( H 1+ ) = . . .

  51. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N = P ( I ∧ H 1+ ) P ( I | H 1+ ) CP � I → H 1+ � P ( H 1+ ) P ( I ) = P ( H 1+ ) = . . . = N − λ ( N − C − 1) N

  52. analysis general probability II P ( I ) TP = P ( H 1 ) P ( I | H 1 ) + P ( H 2+ ) P ( I | H 2+ ) = . . . = λ · C � λ � · 1 + N − λ · ( N − C ) N = P ( I ∧ H 1+ ) P ( I | H 1+ ) CP � I → H 1+ � P ( H 1+ ) P ( I ) = P ( H 1+ ) = . . . = N − λ ( N − C − 1) N good node P (good node i is initiator) = 1 − P ( I | H 1+ ) = λ N < 1 N ≤ 1 N − C − 1 2

Recommend


More recommend