Games for Exchanging Information

33
Games for Exchanging Information Gillat Kol Joint work with Moni Naor

description

Games for Exchanging Information. Gillat Kol Joint work with Moni Naor. Our Goal. Design secret sharing schemes that work assuming players are rational. Talk Plan. Introduction Background Related Work Our Contributions Scheme Construction Impossibility Solution Concept. - PowerPoint PPT Presentation

Transcript of Games for Exchanging Information

Page 1: Games for Exchanging  Information

Games for Exchanging Information

Gillat Kol

Joint work with Moni Naor

Page 2: Games for Exchanging  Information

Our Goal

Design secret sharing schemes that work assuming players are rational

Page 3: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

Page 4: Games for Exchanging  Information

Cryptographic vs. Game Theoretic Settings

Cryptography: Players are either arbitrarily malicious or totally honest.

Game Theory: Players are rational trying to maximize their payoff functions. ui(σ) is i’s payoff when following the protocol σ=(σ1,..,σn).

We assume: Players are rational:

Prefer to learn the secret above all else. Secondly, prefer to learn alone.

Players are computationally unbounded. Communicating via a simultaneous broadcast channel

(SBC) - no rushing.

Page 5: Games for Exchanging  Information

Rational Secret Sharing (RSS)

MetaDef: m-out-of-n RSS scheme. Shares assignment algorithm for the dealer (as in the

usual crypto setting). Game Theoretically stable (e.g., Nash equilibrium)

reconstruction protocol for the players.

Def: σ is a Nash Equilibrium no player can gain by deviating from his strategy, assuming that all the others are following theirs:

i σ’i: ui(σi,σ-i) ≥ ui(σ’i,σ-i) Each player’s strategy is a best response to the

strategies of the others.

Page 6: Games for Exchanging  Information

Is Shamir’s scheme an RSS?

Shamir’s scheme is not RSS. Recall that to reconstruct players reveal their shares. For p=m (p = num of participants): Not Nash

Higher payoff for keeping silence. For p>m: “Unstable” Nash

No player, on its own, can prevent others from learning. Silence is never worse revealing, but sometimes better.

Main Problem: Players deviate in the last round of the protocol, since they no longer fear future punishment.

Solution: Players shouldn't be able to identify the last round. Protocols are unbounded and allow players to learn w.p. 1.

Page 7: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

Page 8: Games for Exchanging  Information

Previous Works

Previous results required one of the followings: The dealer’s involvement in the reconstruction [HT04].

Cryptographic tools [GK06, LT06, ADGH06]. Requires computational assumptions and bounded

players. Achieves only approximated Nash.

Different (stronger) hardware assumptions: Private channels [GK06, ADGH06] + [BGW88].

Requires ≥ 4 players. Envelopes and ballots boxes [LMPS04, LMS05, ILM05].

Solve a more general problem (SFE given any utilities). Achieve stronger solution concepts (coalitions).

Page 9: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

Page 10: Games for Exchanging  Information

Our Contribution

Solution Concept: What is a good RSS scheme? Previous criterion does not rule out all unstable protocols. Previous crypto protocols are susceptible to backward

induction (BI).

Impossibility: There is no “reasonable” Nash RSS with SBC taking shares from finite sets.

Constructing an RSS with SBC and finite shares taken from infinite sets. Satisfies stronger solution concepts (strict Nash, no BI). Unbounded players, No computational assumptions. Can remove the simultaneity assumption and get

approximated Nash.

Page 11: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

Page 12: Games for Exchanging  Information

The Scheme Construction

Present a buggy 2-out-of-2 RSS. Fix it. Analyze it. Generalize to m-out-of-n for all 2≤m≤n. Remove the simultaneity assumption.

Page 13: Games for Exchanging  Information

2-out-of-2 RSS: Dealer’s Algorithm

ℓ1=5

ℓ2=7

L L’

ShortPlayer

LongPlayer

S = {0,..,6}s = 4

2

5

1

3

6

0

4

2

5

1

3

Dealer (s): Uses a parameters (TBD), S is secrets set. Select the shares sizes: ℓ1, ℓ2 = ℓ1+d

where ℓ1,d ~ G() (Geometric distribution).

Select secrets list: random list L of ℓ2 secrets from S s.t. the ℓ1

th secret is s.

Assign shares: choose player randomly, give him L, and the other L’ = L(1,...,ℓ1-1). Players do not know whether their shares

are short or long. Shares are taken from unbounded sets.

Page 14: Games for Exchanging  Information

2-out-of-2 RSS: Player’s Algorithm

Iteration 1

L L’

ShortPlayer

LongPlayer

2

5

1

3

6

0

2

5

1

3

Iteration 2

Iteration 3

Iteration 4

Iteration 5quiet4

S = {0,..,6}s = 4 Player (share):

Broadcast the next secret in your list. Keep silent if your list ended.

If the other broadcasted a false value, abort.

If only a single player broadcasts: the last value broadcasted is s.

Page 15: Games for Exchanging  Information

Bug 1: Identifying the Last Iteration

Problem: The short player identifies the last iteration when his list ends. May broadcast a fictitious secret.

Solution: Divide iterations into stages: #stages in each iteration is chosen ~ G(). Players broadcast only during the last stage. Players get #stages for cells in their list.

The short player does not know #stages of the last iteration.

4 2

8 5

2 1

2 3

6 4

3 6

7 0

Secrets#Stages

ShortPlayer

LongPlayer

4 2

8 5

2 1

2 3

Page 16: Games for Exchanging  Information

Bug 2: Guessing the Secret

Problem: If some secret appears a lot in the list, w.h.p it is the real secret.

Solution: Mask every secret in the list using a random mask Dealer gives each player a share of every mask. Shares of the tth mask are broadcasted by the

players during iteration t-1.

4

4

4

4

4

4

4

L

Page 17: Games for Exchanging  Information

Bug 3: Broadcasting Fictitious Information

Problem: Players may broadcast fictitious information.

Solution: Dealer equip players with authentication information.

Now it works…

Page 18: Games for Exchanging  Information

Strict Nash Equilibrium

Def: σ is a Strict Nash Equilibrium every player looses when deviating from his strategy, assuming that all the others are following theirs:

i σ’i: ui(σi,σ-i) > ui(σ’i,σ-i) A player’s strategy is a strict, unique best response.

Strict Nash Nash Example: Shamir’s reconstruction is not a strict Nash.

Page 19: Games for Exchanging  Information

Protocol Analysis

Recall: Pr[ current iteration is the last ] = . Theorem: For a sufficiently small , the scheme is a

strict Nash with expected number of rounds 1/2.

Proof: By deviating players risk early termination.

must depend on the payoffs. The higher the payoff for learning alone vs. learning

with others, the smaller is.

Page 20: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

Page 21: Games for Exchanging  Information

Revelation Point Theorem: There is no Nash RSS with shares taken

from finite sets without a revelation point (RP). Def (Informal): RP of a reconstruction protocol is a

point its execution for which: Some players do not know the secret. At any point after it, the secret is known to all.

Protocols with RP are “unreasonable”. Players always learn after RP Should not reveal info. Players learn right after RP Someone does reveal info.

Example: Shamir’s reconstruction has RP before the first round.

Strict Nash Nash with no RP

Page 22: Games for Exchanging  Information

Transcripts Trees

A transcript of σ is a possible sequence of messages m = (m1,…,mℓ) broadcasted by the players during rounds 1..ℓ while following σ.

We view transcripts as vertices of a Transcripts Tree. Def: RP of σ is a vertex in σ’s transcript tree that has

children, but no grandchildren.

Page 23: Games for Exchanging  Information

Claim: Children are Correlated

Assume for simplicity that σ allows players to learn together.

Claim: For every transcript p of σ, one of the following holds:

Players always learn after the next round. Players never learn after the next round.

(independently of their random tapes)

Impossible: all learn

pno-onelearns

Page 24: Games for Exchanging  Information

Claim Proof: Hybrid Argument

Proof: Assume that the input is x, and that players learn

given r = (r1,..,rn), but don’t learn given r’ = (r1’,..,rn’). Define the hybrid ri = (r’1,..,r’i,ri+1,..,rn). Hybrid Argument: i s.t. given shares x, all learn

given ri, but no-one learns given ri+1. Players other than i act the same given ri and ri+1

i learns given ri+1 since he learns given ri Contradiction!

Page 25: Games for Exchanging  Information

Theorem Proof: Inductive Argument

Theorem: There is no Nash RSS with shares taken from finite sets without an RP.

Proof: Construct a path leading to the RP. C(m) = Set of possible shares x for

which players do not know s when reaching m.

m0 = empty transcript. Take x1C(m0). m, a descendent of m0, s.t. given x1,

players learn s after m, but not before.

m0

m1

mk

m2

pre

velation

point

x1

x2

xk

Page 26: Games for Exchanging  Information

Let p be m’s parent. If p has no grandchildren, p is an RP. Otherwise, let m1 be a child of p with children. Using the claim: Players learn after m given shares x1

They learn after m1 given x1. C(m0) C(m1) Recall: C(m) = Set of possible shares for which players do

not know s when reaching m. Use the same argument to find m0, m1, m2… s.t. C(m0)

C(m1) C(m2)… Since the shares sets are finite, the sequence is finite.

Theorem Proof: Inductive argument

The finiteness of the shares set is used! ▪

Page 27: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

On Iterated Admissibility On Backward Induction

Page 28: Games for Exchanging  Information

Previous Criterion: Iterated Admissibility (IA)

IA was used as a criterion distinguishing good from bad schemes in [HT04, GK06, LT06, ADGH06].

Def: Strategy σi is (weakly) dominated if there exists a strategy i that is never worse than σi but sometimes strictly better

(1) σ-i: ui(i , σ-i) ≥ ui(σi, σ-i )

(2) σ-i: ui(i, σ-i) > ui(σi, σ-i) Example: Shamir’s reconstruction is dominated by

the silence strategy. Def: A strategies is Iterated Admissible (IA) if it

survives iterated deletion of dominated strategies.

Page 29: Games for Exchanging  Information

IA doesn’t rule out all bad behaviors

No finite strategy is stable The game played is infinite. talk-oncei = Shamir’s reconstruction in the infinite

game. i reveals his share in round 1 and then broadcasts forever.

Theorem: talk-oncei is IA. Proof:

i trying to dominate talk-oncei there is a “savior” σ-i.

Example: For i = silence, σ-i = others keep silent in round 1, and reveal their shares in round 2 iff i talked in round 1.

In general: σ-i waits to see if player i follows talk-oncei, then rewards or punishes him accordingly.

Strict Nash IA Nash

Page 30: Games for Exchanging  Information

Talk Plan

Introduction Background Related Work

Our Contributions Scheme Construction Impossibility Solution Concept

On Iterated Admissibility On Backward Induction

Page 31: Games for Exchanging  Information

Backward Induction Previous crypto solutions [LT06, ADGH06]:

Run the crypto SFE [GMW87] in every iteration. Have small expected running time, but are unbounded.

Observation: Those protocols are essentially bounded by K iterations (K = #of keys for the SFE of iteration 1).

Problem: Backward Induction The BI process: Players deviate in iteration K since it is the

last, causing K-1 to be last. The same holds for K-1,K-2,..,1. BI causes the instability in exponential events to be

amplified. Solution: Should require the protocol to still be stable

after any history. Our protocol satisfies this property! (as is every exact Nash)

Page 32: Games for Exchanging  Information

Concluding Remarks

Game Theory and Cryptography Common areas of interest (e.g. simulating mediators). Different assumptions and models. By combining techniques / ideas we gain new insights.

We look for RSS schemes using SBC. Solution concept is an issue. The infiniteness of the shares sets is a necessary and

sufficient condition for an exact solution.

Page 33: Games for Exchanging  Information

References[ADGH06] Abraham, Dolev, Gonen, and Halpern. Robust Mechanisms for

Rational Secret Sharing and Multiparty Computation. PODC 2006.

[BGW88] Ben-Or, Goldwasser, Wigderson. Completeness Theorems for Non-Cryptographic Fault-Tolerant Distributed Computation STOC 1988.

[GK06] Gordon and Katz. Rational Secret Sharing, Revisited. SCN 2006.

[GMW87] Goldreich, Micali, and Wigderson. How to Play any Mental Game. STOC 1987.

[HT04] Halpern and Teague. Rational Secret Sharing and Multiparty Computation. STOC 2004.

[ILM05] Izmalkov, Micali, and Lepinski. Rational Secure Computation and Ideal Mechanism Design. FOCS 2005.

[LT06] Lysyanskaya and Triandopoulos. Rationality and Adversarial Behavior in Multi-Party Computation. CRYPTO 2006.

[LMPS04] Lepinski, Micali, Peikert, and Shelat. Completely Fair SFE and Coalition-Safe Cheap Talk. PODC 2004.

[LMS05] Lepinski, Micali, and Shelat. Collusion-Free Protocols. STOC 2005.