Secure Certification of Mixed Quantum States Frdric Dupuis, Serge - - PowerPoint PPT Presentation

secure certification of mixed quantum states
SMART_READER_LITE
LIVE PREVIEW

Secure Certification of Mixed Quantum States Frdric Dupuis, Serge - - PowerPoint PPT Presentation

Secure Certification of Mixed Quantum States Frdric Dupuis, Serge Fehr, Philippe Lamontagne and Louis Salvail Quantum state certification H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H


slide-1
SLIDE 1

Secure Certification of Mixed Quantum States

Frédéric Dupuis, Serge Fehr, Philippe Lamontagne and Louis Salvail

slide-2
SLIDE 2

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H

1/8

slide-3
SLIDE 3

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H

1/8

slide-4
SLIDE 4

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H Certification

  • Measure H with {|ψ

ψ|, I − |ψ ψ|}

  • If result is |ψ for every H , then most of the remaining

positions are in state |ψ with overwhelming probability [BF10].

  • The reference state |ψ must be pure.

1/8

slide-5
SLIDE 5

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H Certification

  • Measure H with {|ψ

ψ|, I − |ψ ψ|}

  • If result is |ψ for every H , then most of the remaining

positions are in state |ψ with overwhelming probability [BF10].

  • The reference state |ψ must be pure.

1/8

slide-6
SLIDE 6

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H Certification

  • Measure H with {|ψ

ψ|, I − |ψ ψ|}

  • If result is |ψ for every H , then most of the remaining

positions are in state |ψ with overwhelming probability [BF10].

  • The reference state |ψ must be pure.

1/8

slide-7
SLIDE 7

Quantum state certification

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H Certification

  • Measure H with {|ψ

ψ|, I − |ψ ψ|}

  • If result is |ψ for every H , then most of the remaining

positions are in state |ψ with overwhelming probability [BF10].

  • The reference state |ψ must be pure.

1/8

slide-8
SLIDE 8

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable

2/8

slide-9
SLIDE 9

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

2/8

slide-10
SLIDE 10

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

2/8

slide-11
SLIDE 11

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

2/8

slide-12
SLIDE 12

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

  • For pure states
  • ψsample
  • = |0⊗k

Pr≈1

= ⇒ |ψrest ∈ span{|x : x has less than δn 1s}

2/8

slide-13
SLIDE 13

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

  • For pure states
  • ψsample
  • = |0⊗k

Pr≈1

= ⇒ |ψrest ∈ span{|x : x has less than δn 1s}

2/8

slide-14
SLIDE 14

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

  • For pure states
  • ψsample
  • = |0⊗k

Pr≈1

= ⇒ |ψrest ∈ span{|x : x has less than δn 1s}

2/8

slide-15
SLIDE 15

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

  • For pure states
  • ψsample
  • = |0⊗k

Pr≈1

= ⇒ |ψrest ∈ span{|x : x has less than δn 1s}

  • For some mixed states ϕ,

supp(ϕ⊗n) = H⊗n

2/8

slide-16
SLIDE 16

What about certifying mixed states ?

Usual approach fail Notion of typical subspace not applicable Xsample = 00 . . . 0 Pr≈1 = ⇒ Xrest ∈ {x : x has less than δn 1s}

  • For pure states
  • ψsample
  • = |0⊗k

Pr≈1

= ⇒ |ψrest ∈ span{|x : x has less than δn 1s}

  • For some mixed states ϕ,

supp(ϕ⊗n) = H⊗n No local measurement for a discrete notion of errors for mixed states

2/8

slide-17
SLIDE 17

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register.

3/8

slide-18
SLIDE 18

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

3/8

slide-19
SLIDE 19

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

  • P. Prepare |ϕ⊗n

AR, send An to verifier.

  • V. Choose a random sample, announce it to prover.
  • P. Send R for each position in sample.
  • V. Measure {|ϕ

ϕ|AR, I − |ϕ ϕ|AR} for each joint system AR in sample.

  • V. Accept if no errors, reject otherwise.

3/8

slide-20
SLIDE 20

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

  • P. Prepare |ϕ⊗n

AR, send An to verifier.

  • V. Choose a random sample, announce it to prover.
  • P. Send R for each position in sample.
  • V. Measure {|ϕ

ϕ|AR, I − |ϕ ϕ|AR} for each joint system AR in sample.

  • V. Accept if no errors, reject otherwise.

3/8

slide-21
SLIDE 21

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

  • P. Prepare |ϕ⊗n

AR, send An to verifier.

  • V. Choose a random sample, announce it to prover.
  • P. Send R for each position in sample.
  • V. Measure {|ϕ

ϕ|AR, I − |ϕ ϕ|AR} for each joint system AR in sample.

  • V. Accept if no errors, reject otherwise.

3/8

slide-22
SLIDE 22

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

  • P. Prepare |ϕ⊗n

AR, send An to verifier.

  • V. Choose a random sample, announce it to prover.
  • P. Send R for each position in sample.
  • V. Measure {|ϕ

ϕ|AR, I − |ϕ ϕ|AR} for each joint system AR in sample.

  • V. Accept if no errors, reject otherwise.

3/8

slide-23
SLIDE 23

A mixed state certification protocol

Possible to verify that a qubit is in state ϕ if we have access to its purifying register. Two-player «Game» Verifier wants to certify that his state is close to ϕ⊗n. Prover wants to fool the verifier into thinking he has the right state even though it’s not the case.

  • P. Prepare |ϕ⊗n

AR, send An to verifier.

  • V. Choose a random sample, announce it to prover.
  • P. Send R for each position in sample.
  • V. Measure {|ϕ

ϕ|AR, I − |ϕ ϕ|AR} for each joint system AR in sample.

  • V. Accept if no errors, reject otherwise.

3/8

slide-24
SLIDE 24

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

4/8

slide-25
SLIDE 25

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

4/8

slide-26
SLIDE 26

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

  • 1. Learns sample

4/8

slide-27
SLIDE 27

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

  • 1. Learns sample
  • 2. Measures qubits

4/8

slide-28
SLIDE 28

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

Abort/continue

  • 1. Learns sample
  • 2. Measures qubits
  • 3. Aborts based on result

4/8

slide-29
SLIDE 29

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

Abort/continue

  • 1. Learns sample
  • 2. Measures qubits
  • 3. Aborts based on result

P

  • s

t

  • s

e l e c t i

  • n

4/8

slide-30
SLIDE 30

A few observations about the protocol

Interaction is necessary How can you distinguish |0 0| 2 + |1 1| 2 ⊗n from

≈n/2 times

  • |0|0 . . . |0

≈n/2 times

  • |1|1 . . . |1

Interaction gives more power to prover

P. V.

Abort/continue

  • 1. Learns sample
  • 2. Measures qubits
  • 3. Aborts based on result

P

  • s

t

  • s

e l e c t i

  • n

Example

Prepare

1 √ 2(|00 + |11)⊗n,

measure positions outside of sample, abort if result = |0⊗n−k. Resulting state always |0⊗n−k

4/8

slide-31
SLIDE 31

What can the prover do ?

5/8

slide-32
SLIDE 32

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

5/8

slide-33
SLIDE 33

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

|ϕ⊗n = |ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ

5/8

slide-34
SLIDE 34

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

|ϕ⊗n = |ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ |ψe = |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ|ϕ |ϕ |ϕ|ϕ

5/8

slide-35
SLIDE 35

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

|ϕ⊗n = |ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ|ϕ |ψe = |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ|ϕ |ϕ |ϕ|ϕ ρAnRn =

  • e

pe|ψe ψe|

5/8

slide-36
SLIDE 36

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

· · · |ψe = |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ |ϕ |ϕ|ϕ|ϕ |ϕ |ϕ|ϕ |ϕ |ϕ|ϕ ρAnRn =

  • e

pe|ψe ψe| |ΨAnRnE =

e

√pe|ψeAnRn ⊗ |τeE

5/8

slide-37
SLIDE 37

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

· · · ρAnRn =

  • e

pe|ψe ψe| |ΨAnRnE =

e

√pe|ψeAnRn ⊗ |τeE |ˆ ΨAnRnE = IAn ⊗ MRnE|ΨAnRnE

5/8

slide-38
SLIDE 38

What can the prover do ?

An “undetectable” attack The prover can

  • prepare the honest state, up to a few errors,
  • prepare a mixture/superposition of such states,
  • purify this mixture, and
  • post-select on a measurement outcome.

· · · ρAnRn =

  • e

pe|ψe ψe| |ΨAnRnE =

e

√pe|ψeAnRn ⊗ |τeE |ˆ ΨAnRnE = IAn ⊗ MRnE|ΨAnRnE ideal state

5/8

slide-39
SLIDE 39

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-40
SLIDE 40

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-41
SLIDE 41

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-42
SLIDE 42

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-43
SLIDE 43

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-44
SLIDE 44

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

6/8

slide-45
SLIDE 45

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

Application to Cryptography For any POVM operator E bad of a “bad” outcome, tr

  • E badρAn
  • ≤ pn · tr
  • E badψAn
  • + negl(n)

Bad outcome on real state has negligible probability if tr(E badψAn) is negligible.

6/8

slide-46
SLIDE 46

The mixed state certification Theorem

Main Result For any strategy of the prover, if the verifier accepts, his output state ρAn satisfies ρAn ≤ pn · ψAn + σ where pn is a fixed-degree polynomial in n, ψAn is the reduced

  • perator of an ideal state |ψAnRnE and tr(σ) ≤ negl(n).

Application to Cryptography For any POVM operator E bad of a “bad” outcome, tr

  • E badρAn
  • ≤ pn · tr
  • E badψAn
  • + negl(n)

Bad outcome on real state has negligible probability if tr(E badψAn) is negligible.

6/8

slide-47
SLIDE 47

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ.

7/8

slide-48
SLIDE 48

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ.

7/8

slide-49
SLIDE 49

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ. Corollary Theorem implies security of

  • a local measurement certification protocol for ϕ = I

2,

  • pure state certification [BF10], and
  • a “distributed” pure state certification protocol [DDN14] not

covered by [BF10].

7/8

slide-50
SLIDE 50

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ. Corollary Theorem implies security of

  • a local measurement certification protocol for ϕ = I

2,

  • pure state certification [BF10], and
  • a “distributed” pure state certification protocol [DDN14] not

covered by [BF10].

7/8

slide-51
SLIDE 51

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ. Corollary Theorem implies security of

  • a local measurement certification protocol for ϕ = I

2,

  • pure state certification [BF10], and
  • a “distributed” pure state certification protocol [DDN14] not

covered by [BF10].

7/8

slide-52
SLIDE 52

Generalisations and special cases

Sufficient conditions Invariance under permutations. Equivalent to protocol where verifier permutes his registers with random π and announces π to the prover. Behaves well on “easy” state. The verifier detects any cheating attempt with overwhelming probability on a state of the form σ⊗n for σ distant from reference state ϕ. Corollary Theorem implies security of

  • a local measurement certification protocol for ϕ = I

2,

  • pure state certification [BF10], and
  • a “distributed” pure state certification protocol [DDN14] not

covered by [BF10].

7/8

slide-53
SLIDE 53

Application : secure two-party randomness generation

7/8

slide-54
SLIDE 54

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability.

8/8

slide-55
SLIDE 55

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability.

8/8

slide-56
SLIDE 56

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability.

8/8

slide-57
SLIDE 57

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability. Protocol

  • Alice prepares |Ψ⊗N

AB and sends BN to Bob.

  • Bob certifies that most of his registers are close to I

2.

  • Alice and Bob measure their remaining n registers.

8/8

slide-58
SLIDE 58

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability. Protocol

  • Alice prepares |Ψ⊗N

AB and sends BN to Bob.

  • Bob certifies that most of his registers are close to I

2.

  • Alice and Bob measure their remaining n registers.

8/8

slide-59
SLIDE 59

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability. Protocol

  • Alice prepares |Ψ⊗N

AB and sends BN to Bob.

  • Bob certifies that most of his registers are close to I

2.

  • Alice and Bob measure their remaining n registers.

8/8

slide-60
SLIDE 60

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability. Protocol

  • Alice prepares |Ψ⊗N

AB and sends BN to Bob.

  • Bob certifies that most of his registers are close to I

2.

  • Alice and Bob measure their remaining n registers.

8/8

slide-61
SLIDE 61

Secure Two-Party Randomness Generation

Goal Produce XA, XB ∈ {0, 1}n such that

  • XA = XB if Alice and Bob are both honest,
  • H∞(XA) ≥ (1 − ǫ)n and H∞(XB) ≥ (1 − ǫ)n except with

negligible probability. Protocol

  • Alice prepares |Ψ⊗N

AB and sends BN to Bob.

  • Bob certifies that most of his registers are close to I

2.

  • Alice and Bob measure their remaining n registers.

Our main result ensures that the measurement

  • utcome will have near maximal min-entropy

8/8

slide-62
SLIDE 62

Thank you !

8/8

slide-63
SLIDE 63

Niek J. Bouman and Serge Fehr. Sampling in a quantum population, and applications. In Advances in Cryptology - CRYPTO 2010, 30th Annual Cryptology Conference, volume 6223 of Lecture Notes in Computer Science, pages 724–741. Springer, 2010. Ivan Damgård, Frédéric Dupuis, and Jesper Buus Nielsen. On the orthogonal vector problem and the feasibility of unconditionally secure leakage resilient computation. Cryptology ePrint Archive, Report 2014/282, 2014.

http://eprint.iacr.org/.

8/8