SlideShare a Scribd company logo
RANDOMIZED BYZANTINE GENERALS
by
Michael O.~Rabin*
Presented by
Kishor Datta Gupta
L20421951
COSC 5302
Motivation
• One of the Fundamental Problem in distributed system is
Consensus problem
• A faulty processor can sends inconsistent information
to the other processors. Which will violate the properties
of distributed system
• Influenced by popularity of dining philosophers problem
Leslie Lamport associated story of generals to increase
the attention for solving fault tolerant system
History
• In 1975 Two generals problems (sometimes called the
Chinese Generals Problem) introduced by E. A.
Akkoyunlu, K. Ekanadham, and R. V. Huber .
• In 1978 “Two Generals Paradox” names by Jim Gray
• In 1980 Leslie Lamport published “Reaching
Agreement in the Presence of Faults” (with Marshall
Pease and Robert Shostak) this paper shows This paper
shows that "Byzantine" faults, in which a faulty processor
sends inconsistent information to the other processors,
can defeat any traditional three-processor algorithm
History (cont)
• In April 1982 Leslie Lamport wrote Byzantine Generals
and Transaction Commit Protocols (with Michael
Fischer) and proved that current Byzantine problem will
lead more failure
• In July 1982, Leslie Lamport solved this problem in a
paper written with Marshall Pease and Robert Shostak
and name “Byzantine General problem” established
• 1982 Ben-Or, M., Another advantage of free choice:
Completely asynchronous agreement protocols,
Abstract.
History (cont)
• July 1983 A weaker Byzantine problem Leslie Lamport
• In 1983 Michael O.~Rabin proposed “Random Byzantine
General” solutions
Background
Consensus Problem :
The consensus problem requires agreement among a
number of processes (or agents) for a single data value.
Some of the processes (agents) may fail or be unreliable in
other ways, so consensus protocols must be fault
tolerant or resilient. The processes must somehow put forth
their candidate values, communicate with one another, and
agree on a single consensus value
Consensus Problem(cont)
• Consensus Problem example
Suppose process p1 ,p2, p3 decided that a data value is
X, if there is another p4 decide that value is not X then
the problem create is called Consensus problem.
• Practical Example: mutual exclusion, election,
transactions
Consensus Problem
• Every processor has an input x є X
• Termination: Eventually every non-faulty processor
must decide on a value y.
• Agreement: All decisions by non-faulty processors must
be the same.
• Validity: If all inputs are the same, then the decision of a
non-faulty processor must equal the common input (this
avoids trivial solutions).
Consensus Problem(cont)
Consensus Problem Failure Reasons
• Link Failure
• Processor Failure
• Byzantine fault
Byzantine fault: If one or more processor became
malicious or faulty that they send wrong messages.
Byzantine Fault
Characteristics
• Most difficult
• Most common
Difference between process failure and Byzantine Failure is
in Byzantine Failure process send wrong information while
in process failure processor doesn’t send any thing. That’s
why Byzantine failure is difficult to detect than Processor
failure
Byzantine Failures Show
Byzantine Problem
• “We imagine that several divisions of the Byzantine army
are camped outside an enemy city, each division
commanded by its own general. The generals can
communicate with one another only by messenger. After
observing the enemy, they must decide upon a common
plan of action. However, some of the generals may be
traitors, trying to prevent the loyal generals from
reaching agreement.”----LESLIE LAMPORT, ROBERT
SHOSTAK, and MARSHALL PEASE
General Byzantine
Problem
o Each division of Byzantine army is directed by its own general.
o There are n Generals, some of which are traitors.
o All armies are camped outside enemy castle, observing enemy.
o Communicate with each other by messengers.
o Requirements:
• A: All loyal generals decide upon the same plan of action
• B: A small number of traitors cannot cause the loyal generals to
adopt a bad plan
Note: We do not need to identify the traitors.
Naïve Solution
All generals send message to all other generals . Majority
Result will be taken as decisions
Failure of the solutions:
o Traitors may send different values to different
generals.
o Loyal generals might get conflicting values from
traitors
Reduction
Interactive Consistency Conditions:
o IC1: All loyal lieutenants obey the same order.
o IC2: If the commanding general is loyal, then every loyal
lieutenant obeys the order he sends
Note: If General is loyal, IC2 => IC1.
Example Scenario
Conditions
• 3 generals, 1 traitor among them.
• Message: Attack / Retreat
For LIEUTENANT1 Who is traitor? COMMANDAR or LIEUTENANT2?
In Fig 1 LIEUTENANT1 has to attack to satisfy IC2. Fig 2 LIEUTENANT1
attacks, LIEUTENANT2 retreats. IC1 violated.
So it’s a impossibly situation.
Limit of traitors
Proof by contradiction.
o Assume there is a solution for 3m Albanians with m
traitors.
o Reduce to 3-General problem.
So we can decide that no solutions with fewer than
3m+1 generals can cope with m traitors.
Solution by Oral Message
Assumptions
A1 – Each message that is sent is delivered correctly.
Assures: Traitors cannot interfere with communication as
third party.
A2 – The receiver of a message knows who sent it.
Assures : Traitors cannot send fake messages
A3 – The absence of a message can be detected.
Assures: Traitors cannot interfere by being silent.
Note: Default order to “retreat” for silent traitor
Oral Message Algorithm
• Algorithm OM(0).
• (1) The commander sends his value to every lieutenant.
• (2) Each lieutenant uses the value he receives from the
commander, or uses the value
RETREAT if he receives no value.
Oral Message
Algorithm(cont)
Here m >0
o Commander sends his value to every Lieutenant (vi)
o Each Lieutenant acts as commander for OM(m-1) and sends
vi to the other n-2 lieutenants (or RETREAT)
o For each i, and each j<>i, let vj be the value lieutenant i
receives from lieutenant j in step (2) using OM(m-1).
Lieutenant i uses the value majority (v1, …, vn-1).
Explanation When
Lieutenant is traitor
When Commander is
traitor
Issue With Oral Message
• Need 3f+1 nodes to tolerate f failures which is expensive
• Difficult because traitors can lie about what others said
Time Complexity O(n^m)
Signed Message Solution
New Assumptions:
(a) A loyal general's signature cannot be forged, and any
alteration of the contents of his signed messages can be
detected.
(b) Anyone can verify the authenticity of a general's
signature.
Step:
• Each lieutenant maintains a set V of properly signed
orders received so far.
• The commander sends a signed order to lieutenants
Signed Message
Algorithm
• A lieutenant receives an order from someone (either
from commander or other lieutenants),
o Verifies authenticity and puts it in V.
o If there are less than m distinct signatures on the order
• Augments orders with signature
• Relays messages to lieutenants who have not seen the order.
• When lieutenant receives no new messages, and use
choice(V) as the desired action.
Explanation
Advantage and Issue
Advantage
• Need f+2 nodes
• Easier because traitors can only lie about other traitors
Issue:
• communication overheads
• Signatures
What if not all generals can reach all other generals
directly?
Missing Path
if the communication graph is 3m-regular, and there are at most m traitors,
the problem can still be solved.
Reliability
Achieving reliability in the face of arbitrary malfunctioning is
a difficult problem, and its solution seems to be inherently
expensive
To avoid expense it is often assumed that a computer may
fail to respond but will never respond incorrectly.
Asynchronous System
No algorithm can guarantee to reach consensus in an
asynchronous system, even with one process crash failure
RANDOMIZED BYZANTINE GENERALS
Based on:
• n processes with up to t faulty ones, t +1computing
phases are necessary for reaching Byzantine
Agreement.
• BG problem has no solution in the asynchronous case.
N=Number of process, T=faulty , G = Generals
Why Randomized
• Certain to achieve Byzantine Agreement
• The expected number of rounds required to do so is 4,
independent of N and T.
• The total expected number of messages =c*n*t, where
c = small constant
• In some case employs a fixed number R of rounds. To
reach Byzantine agreement,. but there is a probability
2^-R of error.
Stage
1. Basic Concepts
2. Authentication
3. Lottery
4. The Agreement Protocol
5. Correctness
6. Bounded Expected Time & Errorless Solutions
Basic Concept
Assumptions:
• Every Gi can directly exchange messages with every
other Gj. Where i,j ϵ n
• Every Gi agreed on a common value in the message
• If one Gi proved faulty for not having the common value
it will remain faulty even it later have the common value
Basic Concept
• If all the proper processes have the same initial message
then the system will be called proper, otherwise faulty.
• The processes reach agreement on the common
message by exchanging information
• We assume that each process Gi has a local phase-
clock p(i), and that pi assigns p (i) := p (i) + 1 at the end
of each phase.
Authentication
• A public directory containing for each participant B a
public key' KB.
• When the participant B needs to authenticate a
message M, he employs a secret key DB to compute
another message 0B(M) =N.
• Every other user can recover M form N by use of KB,
Note: The public-key directory is part of the data in each Pi,
and must be incorporated by a Non-faulty "dealer" at ,the
creation of the processes Gi
LOTTERY
• Algorithm employ a lottery procedure by which the
proper Gi's can agree on a randomly chosen s =0 ,I.
• Using Shamirs A method for obtaining digital signatures
and public-key cryptosystems,
• The lottery procedure admits a parameter m<N, so that
Lottery(m) is the m-th lottery round.
• End of the execution of Lottery(m) by the proper Gi, at
least all the proper processes will share Sm.
THE AGREEMENT
PROTOCOL(BAP)
Structure:
Polling : Gi polls the other processes on their value of the
message.
Lottery: Gi decide common random value
Decision: using s Gi determines whether to adopt the
plurality candidate version of the message obtained
through Polling or his current version of the message
These 3 happen in R time. R = desired reliability
BAP(Polling)
BAP(Lottery)
BAP(Decision)
BAP
Correctness
If Initial message of every proper process = M,
Proper System: End of BAP we shall have message (i)=
M for all proper Gi
Faulty System: With probability at ,least 2^-R , all proper
processes Gi will have the same value for message(i).
Note: Probability for not reaching agreement in R rounds is
2^-R
Bounded expected time
As long as t <n/10 we can reach an agreement ignoring
n and t
It requires 0(n^2) message
Using a wakeup call it could be reduces to 0(nt)
Advantage
• Both applied for Synchronized and asynchronous
version
• Better than Ben-or tossing approach
• With some condition it is a robust solution for the
distributed commit problem
• It is more simpler than other existing approach
Disadvantage
• Synchronized system simplify the algorithm but also
arises issue for coordinated agreement
• For Byzanatine Consensus problem also need to
implement wakeup protocol
• It only resilient under appropriate conditions
Practical Implementations
• For bitcoin peer to peer this solutions could be
implemented
• For distributed commit problem this solution can be
implemented and some cases its implemented
Conclusion
Randomize solution has some upper hand also has some
limit. However in 1984 this solutions creates such effect
which later used to implement various extended solutions
for Byzantine agreement
References
• [1] Leslie Lamport, Robert Shostak and Marshall Pease.
TheByzantine Generals Problem. ACM Transactions
onProgramming Languages and Systems, 4(3): 382-401,
July 1982.
• [2] “AWS S3 Availability Event”,
http://status.aws.amazon.com/s3-20080720.html .
Amazon Rec. 10/2013
• https://en.wikipedia.org/wiki/Consensus_(computer_scie
nce)
• http://marknelson.us/2007/07/23/byzantine/
References
• http://lamport.azurewebsites.net/pubs/pubs.html#implem
entation
• http://www.cs.cornell.edu/courses/cs6410/2009fa/lecture
s/24-byzantine.pdf
• http://www.di.univaq.it/~proietti/slide_algdist2010/5-
Consensus.ppt
• http://lamport.azurewebsites.net/pubs/weak-byz.pdf
• http://cs-www.bu.edu/faculty/homer/537/talks/Sachin-
byzagree-lecture.pdf
References
• http://www.cs.sfu.ca/CourseCentral/401/qshi1/slides/cmp
t_401_byzantine_generals.ppt
• http://webcourse.cs.technion.ac.il/236357/Winter2007-
2008/ho/WCFiles/Anna-
The%20Byzantine%20Generals%20Problem.ppt
• http://www.cs.toronto.edu/~demke/469F.07/Lectures/Lect
ure18.ppt
• http://www-inst.eecs.berkeley.edu/~cs162/fa12/hand-
outs/Original_Byzantine.pdf

More Related Content

Similar to Randomized Byzantine Problem by Rabin

Agreement Protocols, distributed File Systems, Distributed Shared Memory
Agreement Protocols, distributed File Systems, Distributed Shared MemoryAgreement Protocols, distributed File Systems, Distributed Shared Memory
Agreement Protocols, distributed File Systems, Distributed Shared Memory
SHIKHA GAUTAM
 
Fault tolerance-omer-rana
Fault tolerance-omer-ranaFault tolerance-omer-rana
Fault tolerance-omer-rana
Gangadhar Thippannagari
 
Aggrement protocols
Aggrement protocolsAggrement protocols
Aggrement protocols
Mayank Jain
 
B21DA0201_02.ppt
B21DA0201_02.pptB21DA0201_02.ppt
B21DA0201_02.ppt
DrPreethiD1
 
Economics of Decentalized Currency Systems
Economics of Decentalized Currency SystemsEconomics of Decentalized Currency Systems
Economics of Decentalized Currency Systems
Ernie Teo
 
Distributed Computing On Topics of: Leader election + Byzantine algorithms)
Distributed Computing On Topics of: Leader election + Byzantine algorithms)Distributed Computing On Topics of: Leader election + Byzantine algorithms)
Distributed Computing On Topics of: Leader election + Byzantine algorithms)
Osama Askoura
 
Agreement protocol
Agreement protocolAgreement protocol
Agreement protocol
Saif Ali Khan
 
Hyperledger Consensus Algorithms
Hyperledger Consensus AlgorithmsHyperledger Consensus Algorithms
Hyperledger Consensus Algorithms
MabelOza12
 
Error detection and correction
Error detection and correctionError detection and correction
Error detection and correction
Abdul Razaq
 
computer networks Error Detection Methods.pdf
computer networks Error Detection Methods.pdfcomputer networks Error Detection Methods.pdf
computer networks Error Detection Methods.pdf
Balasubramanian699229
 
FastBFT
FastBFTFastBFT
FastBFT
YongraeJo
 
computer Networks Error Detection and Correction.ppt
computer Networks Error Detection and Correction.pptcomputer Networks Error Detection and Correction.ppt
computer Networks Error Detection and Correction.ppt
Jayaprasanna4
 
TorCoin_slides
TorCoin_slidesTorCoin_slides
TorCoin_slides
Miles Richardson
 
Real time-embedded-system-lec-06
Real time-embedded-system-lec-06Real time-embedded-system-lec-06
Real time-embedded-system-lec-06
University of Computer Science and Technology
 
Flexible Symmetric Global Snapshot
Flexible Symmetric Global Snapshot Flexible Symmetric Global Snapshot
Flexible Symmetric Global Snapshot
Ashutosh Jaiswal
 
mod-3-cn-ppt (1).pdf bca bsc btech notes
mod-3-cn-ppt (1).pdf bca bsc btech notesmod-3-cn-ppt (1).pdf bca bsc btech notes
mod-3-cn-ppt (1).pdf bca bsc btech notes
anandhakrishnankv
 
Source coding
Source coding Source coding
Source coding
Shankar Gangaju
 
1. Consensus and agreement algorithms - Introduction.pdf
1. Consensus and agreement algorithms - Introduction.pdf1. Consensus and agreement algorithms - Introduction.pdf
1. Consensus and agreement algorithms - Introduction.pdf
AzmiNizar1
 
error detection.ppt
error detection.ppterror detection.ppt
error detection.ppt
Jayaprasanna4
 
01 what is blockchain
01 what is blockchain01 what is blockchain
01 what is blockchain
BastianBlankenburg
 

Similar to Randomized Byzantine Problem by Rabin (20)

Agreement Protocols, distributed File Systems, Distributed Shared Memory
Agreement Protocols, distributed File Systems, Distributed Shared MemoryAgreement Protocols, distributed File Systems, Distributed Shared Memory
Agreement Protocols, distributed File Systems, Distributed Shared Memory
 
Fault tolerance-omer-rana
Fault tolerance-omer-ranaFault tolerance-omer-rana
Fault tolerance-omer-rana
 
Aggrement protocols
Aggrement protocolsAggrement protocols
Aggrement protocols
 
B21DA0201_02.ppt
B21DA0201_02.pptB21DA0201_02.ppt
B21DA0201_02.ppt
 
Economics of Decentalized Currency Systems
Economics of Decentalized Currency SystemsEconomics of Decentalized Currency Systems
Economics of Decentalized Currency Systems
 
Distributed Computing On Topics of: Leader election + Byzantine algorithms)
Distributed Computing On Topics of: Leader election + Byzantine algorithms)Distributed Computing On Topics of: Leader election + Byzantine algorithms)
Distributed Computing On Topics of: Leader election + Byzantine algorithms)
 
Agreement protocol
Agreement protocolAgreement protocol
Agreement protocol
 
Hyperledger Consensus Algorithms
Hyperledger Consensus AlgorithmsHyperledger Consensus Algorithms
Hyperledger Consensus Algorithms
 
Error detection and correction
Error detection and correctionError detection and correction
Error detection and correction
 
computer networks Error Detection Methods.pdf
computer networks Error Detection Methods.pdfcomputer networks Error Detection Methods.pdf
computer networks Error Detection Methods.pdf
 
FastBFT
FastBFTFastBFT
FastBFT
 
computer Networks Error Detection and Correction.ppt
computer Networks Error Detection and Correction.pptcomputer Networks Error Detection and Correction.ppt
computer Networks Error Detection and Correction.ppt
 
TorCoin_slides
TorCoin_slidesTorCoin_slides
TorCoin_slides
 
Real time-embedded-system-lec-06
Real time-embedded-system-lec-06Real time-embedded-system-lec-06
Real time-embedded-system-lec-06
 
Flexible Symmetric Global Snapshot
Flexible Symmetric Global Snapshot Flexible Symmetric Global Snapshot
Flexible Symmetric Global Snapshot
 
mod-3-cn-ppt (1).pdf bca bsc btech notes
mod-3-cn-ppt (1).pdf bca bsc btech notesmod-3-cn-ppt (1).pdf bca bsc btech notes
mod-3-cn-ppt (1).pdf bca bsc btech notes
 
Source coding
Source coding Source coding
Source coding
 
1. Consensus and agreement algorithms - Introduction.pdf
1. Consensus and agreement algorithms - Introduction.pdf1. Consensus and agreement algorithms - Introduction.pdf
1. Consensus and agreement algorithms - Introduction.pdf
 
error detection.ppt
error detection.ppterror detection.ppt
error detection.ppt
 
01 what is blockchain
01 what is blockchain01 what is blockchain
01 what is blockchain
 

More from Kishor Datta Gupta

GAN introduction.pptx
GAN introduction.pptxGAN introduction.pptx
GAN introduction.pptx
Kishor Datta Gupta
 
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
Kishor Datta Gupta
 
A safer approach to build recommendation systems on unidentifiable data
A safer approach to build recommendation systems on unidentifiable dataA safer approach to build recommendation systems on unidentifiable data
A safer approach to build recommendation systems on unidentifiable data
Kishor Datta Gupta
 
Adversarial Attacks and Defense
Adversarial Attacks and DefenseAdversarial Attacks and Defense
Adversarial Attacks and Defense
Kishor Datta Gupta
 
Who is responsible for adversarial defense
Who is responsible for adversarial defenseWho is responsible for adversarial defense
Who is responsible for adversarial defense
Kishor Datta Gupta
 
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
Kishor Datta Gupta
 
Zero shot learning
Zero shot learning Zero shot learning
Zero shot learning
Kishor Datta Gupta
 
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
Kishor Datta Gupta
 
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Kishor Datta Gupta
 
Machine learning in computer security
Machine learning in computer securityMachine learning in computer security
Machine learning in computer security
Kishor Datta Gupta
 
Policy Based reinforcement Learning for time series Anomaly detection
Policy Based reinforcement Learning for time series Anomaly detectionPolicy Based reinforcement Learning for time series Anomaly detection
Policy Based reinforcement Learning for time series Anomaly detection
Kishor Datta Gupta
 
Cyber intrusion
Cyber intrusionCyber intrusion
Cyber intrusion
Kishor Datta Gupta
 
understanding the pandemic through mining covid news using natural language p...
understanding the pandemic through mining covid news using natural language p...understanding the pandemic through mining covid news using natural language p...
understanding the pandemic through mining covid news using natural language p...
Kishor Datta Gupta
 
Different representation space for MNIST digit
Different representation space for MNIST digitDifferent representation space for MNIST digit
Different representation space for MNIST digit
Kishor Datta Gupta
 
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui..."Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
Kishor Datta Gupta
 
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
Kishor Datta Gupta
 
Adversarial Input Detection Using Image Processing Techniques (IPT)
Adversarial Input Detection Using Image Processing Techniques (IPT)Adversarial Input Detection Using Image Processing Techniques (IPT)
Adversarial Input Detection Using Image Processing Techniques (IPT)
Kishor Datta Gupta
 
Clustering report
Clustering reportClustering report
Clustering report
Kishor Datta Gupta
 
Basic digital image concept
Basic digital image conceptBasic digital image concept
Basic digital image concept
Kishor Datta Gupta
 
An empirical study on algorithmic bias (aiml compsac2020)
An empirical study on algorithmic bias (aiml compsac2020)An empirical study on algorithmic bias (aiml compsac2020)
An empirical study on algorithmic bias (aiml compsac2020)
Kishor Datta Gupta
 

More from Kishor Datta Gupta (20)

GAN introduction.pptx
GAN introduction.pptxGAN introduction.pptx
GAN introduction.pptx
 
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
Interpretable Learning Model for Lower Dimensional Feature Space: A Case stud...
 
A safer approach to build recommendation systems on unidentifiable data
A safer approach to build recommendation systems on unidentifiable dataA safer approach to build recommendation systems on unidentifiable data
A safer approach to build recommendation systems on unidentifiable data
 
Adversarial Attacks and Defense
Adversarial Attacks and DefenseAdversarial Attacks and Defense
Adversarial Attacks and Defense
 
Who is responsible for adversarial defense
Who is responsible for adversarial defenseWho is responsible for adversarial defense
Who is responsible for adversarial defense
 
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
Robust Filtering Schemes for Machine Learning Systems to Defend Adversarial A...
 
Zero shot learning
Zero shot learning Zero shot learning
Zero shot learning
 
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
Using Negative Detectors for Identifying Adversarial Data Manipulation in Mac...
 
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
 
Machine learning in computer security
Machine learning in computer securityMachine learning in computer security
Machine learning in computer security
 
Policy Based reinforcement Learning for time series Anomaly detection
Policy Based reinforcement Learning for time series Anomaly detectionPolicy Based reinforcement Learning for time series Anomaly detection
Policy Based reinforcement Learning for time series Anomaly detection
 
Cyber intrusion
Cyber intrusionCyber intrusion
Cyber intrusion
 
understanding the pandemic through mining covid news using natural language p...
understanding the pandemic through mining covid news using natural language p...understanding the pandemic through mining covid news using natural language p...
understanding the pandemic through mining covid news using natural language p...
 
Different representation space for MNIST digit
Different representation space for MNIST digitDifferent representation space for MNIST digit
Different representation space for MNIST digit
 
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui..."Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
"Can NLP techniques be utilized as a reliable tool for medical science?" -Bui...
 
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
Applicability issues of Evasion-Based Adversarial Attacks and Mitigation Tech...
 
Adversarial Input Detection Using Image Processing Techniques (IPT)
Adversarial Input Detection Using Image Processing Techniques (IPT)Adversarial Input Detection Using Image Processing Techniques (IPT)
Adversarial Input Detection Using Image Processing Techniques (IPT)
 
Clustering report
Clustering reportClustering report
Clustering report
 
Basic digital image concept
Basic digital image conceptBasic digital image concept
Basic digital image concept
 
An empirical study on algorithmic bias (aiml compsac2020)
An empirical study on algorithmic bias (aiml compsac2020)An empirical study on algorithmic bias (aiml compsac2020)
An empirical study on algorithmic bias (aiml compsac2020)
 

Recently uploaded

The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptxThe use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
MAGOTI ERNEST
 
Phenomics assisted breeding in crop improvement
Phenomics assisted breeding in crop improvementPhenomics assisted breeding in crop improvement
Phenomics assisted breeding in crop improvement
IshaGoswami9
 
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốtmô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
HongcNguyn6
 
molar-distalization in orthodontics-seminar.pptx
molar-distalization in orthodontics-seminar.pptxmolar-distalization in orthodontics-seminar.pptx
molar-distalization in orthodontics-seminar.pptx
Anagha Prasad
 
20240520 Planning a Circuit Simulator in JavaScript.pptx
20240520 Planning a Circuit Simulator in JavaScript.pptx20240520 Planning a Circuit Simulator in JavaScript.pptx
20240520 Planning a Circuit Simulator in JavaScript.pptx
Sharon Liu
 
Equivariant neural networks and representation theory
Equivariant neural networks and representation theoryEquivariant neural networks and representation theory
Equivariant neural networks and representation theory
Daniel Tubbenhauer
 
Eukaryotic Transcription Presentation.pptx
Eukaryotic Transcription Presentation.pptxEukaryotic Transcription Presentation.pptx
Eukaryotic Transcription Presentation.pptx
RitabrataSarkar3
 
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
AbdullaAlAsif1
 
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdfTopic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
TinyAnderson
 
Direct Seeded Rice - Climate Smart Agriculture
Direct Seeded Rice - Climate Smart AgricultureDirect Seeded Rice - Climate Smart Agriculture
Direct Seeded Rice - Climate Smart Agriculture
International Food Policy Research Institute- South Asia Office
 
THEMATIC APPERCEPTION TEST(TAT) cognitive abilities, creativity, and critic...
THEMATIC  APPERCEPTION  TEST(TAT) cognitive abilities, creativity, and critic...THEMATIC  APPERCEPTION  TEST(TAT) cognitive abilities, creativity, and critic...
THEMATIC APPERCEPTION TEST(TAT) cognitive abilities, creativity, and critic...
Abdul Wali Khan University Mardan,kP,Pakistan
 
aziz sancar nobel prize winner: from mardin to nobel
aziz sancar nobel prize winner: from mardin to nobelaziz sancar nobel prize winner: from mardin to nobel
aziz sancar nobel prize winner: from mardin to nobel
İsa Badur
 
Sciences of Europe journal No 142 (2024)
Sciences of Europe journal No 142 (2024)Sciences of Europe journal No 142 (2024)
Sciences of Europe journal No 142 (2024)
Sciences of Europe
 
Authoring a personal GPT for your research and practice: How we created the Q...
Authoring a personal GPT for your research and practice: How we created the Q...Authoring a personal GPT for your research and practice: How we created the Q...
Authoring a personal GPT for your research and practice: How we created the Q...
Leonel Morgado
 
Thornton ESPP slides UK WW Network 4_6_24.pdf
Thornton ESPP slides UK WW Network 4_6_24.pdfThornton ESPP slides UK WW Network 4_6_24.pdf
Thornton ESPP slides UK WW Network 4_6_24.pdf
European Sustainable Phosphorus Platform
 
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
Sérgio Sacani
 
GBSN - Biochemistry (Unit 6) Chemistry of Proteins
GBSN - Biochemistry (Unit 6) Chemistry of ProteinsGBSN - Biochemistry (Unit 6) Chemistry of Proteins
GBSN - Biochemistry (Unit 6) Chemistry of Proteins
Areesha Ahmad
 
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
vluwdy49
 
Bob Reedy - Nitrate in Texas Groundwater.pdf
Bob Reedy - Nitrate in Texas Groundwater.pdfBob Reedy - Nitrate in Texas Groundwater.pdf
Bob Reedy - Nitrate in Texas Groundwater.pdf
Texas Alliance of Groundwater Districts
 
8.Isolation of pure cultures and preservation of cultures.pdf
8.Isolation of pure cultures and preservation of cultures.pdf8.Isolation of pure cultures and preservation of cultures.pdf
8.Isolation of pure cultures and preservation of cultures.pdf
by6843629
 

Recently uploaded (20)

The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptxThe use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
The use of Nauplii and metanauplii artemia in aquaculture (brine shrimp).pptx
 
Phenomics assisted breeding in crop improvement
Phenomics assisted breeding in crop improvementPhenomics assisted breeding in crop improvement
Phenomics assisted breeding in crop improvement
 
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốtmô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
mô tả các thí nghiệm về đánh giá tác động dòng khí hóa sau đốt
 
molar-distalization in orthodontics-seminar.pptx
molar-distalization in orthodontics-seminar.pptxmolar-distalization in orthodontics-seminar.pptx
molar-distalization in orthodontics-seminar.pptx
 
20240520 Planning a Circuit Simulator in JavaScript.pptx
20240520 Planning a Circuit Simulator in JavaScript.pptx20240520 Planning a Circuit Simulator in JavaScript.pptx
20240520 Planning a Circuit Simulator in JavaScript.pptx
 
Equivariant neural networks and representation theory
Equivariant neural networks and representation theoryEquivariant neural networks and representation theory
Equivariant neural networks and representation theory
 
Eukaryotic Transcription Presentation.pptx
Eukaryotic Transcription Presentation.pptxEukaryotic Transcription Presentation.pptx
Eukaryotic Transcription Presentation.pptx
 
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
Unlocking the mysteries of reproduction: Exploring fecundity and gonadosomati...
 
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdfTopic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
Topic: SICKLE CELL DISEASE IN CHILDREN-3.pdf
 
Direct Seeded Rice - Climate Smart Agriculture
Direct Seeded Rice - Climate Smart AgricultureDirect Seeded Rice - Climate Smart Agriculture
Direct Seeded Rice - Climate Smart Agriculture
 
THEMATIC APPERCEPTION TEST(TAT) cognitive abilities, creativity, and critic...
THEMATIC  APPERCEPTION  TEST(TAT) cognitive abilities, creativity, and critic...THEMATIC  APPERCEPTION  TEST(TAT) cognitive abilities, creativity, and critic...
THEMATIC APPERCEPTION TEST(TAT) cognitive abilities, creativity, and critic...
 
aziz sancar nobel prize winner: from mardin to nobel
aziz sancar nobel prize winner: from mardin to nobelaziz sancar nobel prize winner: from mardin to nobel
aziz sancar nobel prize winner: from mardin to nobel
 
Sciences of Europe journal No 142 (2024)
Sciences of Europe journal No 142 (2024)Sciences of Europe journal No 142 (2024)
Sciences of Europe journal No 142 (2024)
 
Authoring a personal GPT for your research and practice: How we created the Q...
Authoring a personal GPT for your research and practice: How we created the Q...Authoring a personal GPT for your research and practice: How we created the Q...
Authoring a personal GPT for your research and practice: How we created the Q...
 
Thornton ESPP slides UK WW Network 4_6_24.pdf
Thornton ESPP slides UK WW Network 4_6_24.pdfThornton ESPP slides UK WW Network 4_6_24.pdf
Thornton ESPP slides UK WW Network 4_6_24.pdf
 
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
EWOCS-I: The catalog of X-ray sources in Westerlund 1 from the Extended Weste...
 
GBSN - Biochemistry (Unit 6) Chemistry of Proteins
GBSN - Biochemistry (Unit 6) Chemistry of ProteinsGBSN - Biochemistry (Unit 6) Chemistry of Proteins
GBSN - Biochemistry (Unit 6) Chemistry of Proteins
 
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
在线办理(salfor毕业证书)索尔福德大学毕业证毕业完成信一模一样
 
Bob Reedy - Nitrate in Texas Groundwater.pdf
Bob Reedy - Nitrate in Texas Groundwater.pdfBob Reedy - Nitrate in Texas Groundwater.pdf
Bob Reedy - Nitrate in Texas Groundwater.pdf
 
8.Isolation of pure cultures and preservation of cultures.pdf
8.Isolation of pure cultures and preservation of cultures.pdf8.Isolation of pure cultures and preservation of cultures.pdf
8.Isolation of pure cultures and preservation of cultures.pdf
 

Randomized Byzantine Problem by Rabin

  • 1. RANDOMIZED BYZANTINE GENERALS by Michael O.~Rabin* Presented by Kishor Datta Gupta L20421951 COSC 5302
  • 2. Motivation • One of the Fundamental Problem in distributed system is Consensus problem • A faulty processor can sends inconsistent information to the other processors. Which will violate the properties of distributed system • Influenced by popularity of dining philosophers problem Leslie Lamport associated story of generals to increase the attention for solving fault tolerant system
  • 3. History • In 1975 Two generals problems (sometimes called the Chinese Generals Problem) introduced by E. A. Akkoyunlu, K. Ekanadham, and R. V. Huber . • In 1978 “Two Generals Paradox” names by Jim Gray • In 1980 Leslie Lamport published “Reaching Agreement in the Presence of Faults” (with Marshall Pease and Robert Shostak) this paper shows This paper shows that "Byzantine" faults, in which a faulty processor sends inconsistent information to the other processors, can defeat any traditional three-processor algorithm
  • 4. History (cont) • In April 1982 Leslie Lamport wrote Byzantine Generals and Transaction Commit Protocols (with Michael Fischer) and proved that current Byzantine problem will lead more failure • In July 1982, Leslie Lamport solved this problem in a paper written with Marshall Pease and Robert Shostak and name “Byzantine General problem” established • 1982 Ben-Or, M., Another advantage of free choice: Completely asynchronous agreement protocols, Abstract.
  • 5. History (cont) • July 1983 A weaker Byzantine problem Leslie Lamport • In 1983 Michael O.~Rabin proposed “Random Byzantine General” solutions
  • 6. Background Consensus Problem : The consensus problem requires agreement among a number of processes (or agents) for a single data value. Some of the processes (agents) may fail or be unreliable in other ways, so consensus protocols must be fault tolerant or resilient. The processes must somehow put forth their candidate values, communicate with one another, and agree on a single consensus value
  • 7. Consensus Problem(cont) • Consensus Problem example Suppose process p1 ,p2, p3 decided that a data value is X, if there is another p4 decide that value is not X then the problem create is called Consensus problem. • Practical Example: mutual exclusion, election, transactions
  • 8. Consensus Problem • Every processor has an input x є X • Termination: Eventually every non-faulty processor must decide on a value y. • Agreement: All decisions by non-faulty processors must be the same. • Validity: If all inputs are the same, then the decision of a non-faulty processor must equal the common input (this avoids trivial solutions).
  • 9. Consensus Problem(cont) Consensus Problem Failure Reasons • Link Failure • Processor Failure • Byzantine fault Byzantine fault: If one or more processor became malicious or faulty that they send wrong messages.
  • 10. Byzantine Fault Characteristics • Most difficult • Most common Difference between process failure and Byzantine Failure is in Byzantine Failure process send wrong information while in process failure processor doesn’t send any thing. That’s why Byzantine failure is difficult to detect than Processor failure
  • 12. Byzantine Problem • “We imagine that several divisions of the Byzantine army are camped outside an enemy city, each division commanded by its own general. The generals can communicate with one another only by messenger. After observing the enemy, they must decide upon a common plan of action. However, some of the generals may be traitors, trying to prevent the loyal generals from reaching agreement.”----LESLIE LAMPORT, ROBERT SHOSTAK, and MARSHALL PEASE
  • 13. General Byzantine Problem o Each division of Byzantine army is directed by its own general. o There are n Generals, some of which are traitors. o All armies are camped outside enemy castle, observing enemy. o Communicate with each other by messengers. o Requirements: • A: All loyal generals decide upon the same plan of action • B: A small number of traitors cannot cause the loyal generals to adopt a bad plan Note: We do not need to identify the traitors.
  • 14. Naïve Solution All generals send message to all other generals . Majority Result will be taken as decisions Failure of the solutions: o Traitors may send different values to different generals. o Loyal generals might get conflicting values from traitors
  • 15. Reduction Interactive Consistency Conditions: o IC1: All loyal lieutenants obey the same order. o IC2: If the commanding general is loyal, then every loyal lieutenant obeys the order he sends Note: If General is loyal, IC2 => IC1.
  • 16. Example Scenario Conditions • 3 generals, 1 traitor among them. • Message: Attack / Retreat For LIEUTENANT1 Who is traitor? COMMANDAR or LIEUTENANT2? In Fig 1 LIEUTENANT1 has to attack to satisfy IC2. Fig 2 LIEUTENANT1 attacks, LIEUTENANT2 retreats. IC1 violated. So it’s a impossibly situation.
  • 17. Limit of traitors Proof by contradiction. o Assume there is a solution for 3m Albanians with m traitors. o Reduce to 3-General problem. So we can decide that no solutions with fewer than 3m+1 generals can cope with m traitors.
  • 18. Solution by Oral Message Assumptions A1 – Each message that is sent is delivered correctly. Assures: Traitors cannot interfere with communication as third party. A2 – The receiver of a message knows who sent it. Assures : Traitors cannot send fake messages A3 – The absence of a message can be detected. Assures: Traitors cannot interfere by being silent. Note: Default order to “retreat” for silent traitor
  • 19. Oral Message Algorithm • Algorithm OM(0). • (1) The commander sends his value to every lieutenant. • (2) Each lieutenant uses the value he receives from the commander, or uses the value RETREAT if he receives no value.
  • 20. Oral Message Algorithm(cont) Here m >0 o Commander sends his value to every Lieutenant (vi) o Each Lieutenant acts as commander for OM(m-1) and sends vi to the other n-2 lieutenants (or RETREAT) o For each i, and each j<>i, let vj be the value lieutenant i receives from lieutenant j in step (2) using OM(m-1). Lieutenant i uses the value majority (v1, …, vn-1).
  • 23. Issue With Oral Message • Need 3f+1 nodes to tolerate f failures which is expensive • Difficult because traitors can lie about what others said Time Complexity O(n^m)
  • 24. Signed Message Solution New Assumptions: (a) A loyal general's signature cannot be forged, and any alteration of the contents of his signed messages can be detected. (b) Anyone can verify the authenticity of a general's signature. Step: • Each lieutenant maintains a set V of properly signed orders received so far. • The commander sends a signed order to lieutenants
  • 25. Signed Message Algorithm • A lieutenant receives an order from someone (either from commander or other lieutenants), o Verifies authenticity and puts it in V. o If there are less than m distinct signatures on the order • Augments orders with signature • Relays messages to lieutenants who have not seen the order. • When lieutenant receives no new messages, and use choice(V) as the desired action.
  • 27. Advantage and Issue Advantage • Need f+2 nodes • Easier because traitors can only lie about other traitors Issue: • communication overheads • Signatures What if not all generals can reach all other generals directly?
  • 28. Missing Path if the communication graph is 3m-regular, and there are at most m traitors, the problem can still be solved.
  • 29. Reliability Achieving reliability in the face of arbitrary malfunctioning is a difficult problem, and its solution seems to be inherently expensive To avoid expense it is often assumed that a computer may fail to respond but will never respond incorrectly.
  • 30. Asynchronous System No algorithm can guarantee to reach consensus in an asynchronous system, even with one process crash failure
  • 31. RANDOMIZED BYZANTINE GENERALS Based on: • n processes with up to t faulty ones, t +1computing phases are necessary for reaching Byzantine Agreement. • BG problem has no solution in the asynchronous case. N=Number of process, T=faulty , G = Generals
  • 32. Why Randomized • Certain to achieve Byzantine Agreement • The expected number of rounds required to do so is 4, independent of N and T. • The total expected number of messages =c*n*t, where c = small constant • In some case employs a fixed number R of rounds. To reach Byzantine agreement,. but there is a probability 2^-R of error.
  • 33. Stage 1. Basic Concepts 2. Authentication 3. Lottery 4. The Agreement Protocol 5. Correctness 6. Bounded Expected Time & Errorless Solutions
  • 34. Basic Concept Assumptions: • Every Gi can directly exchange messages with every other Gj. Where i,j ϵ n • Every Gi agreed on a common value in the message • If one Gi proved faulty for not having the common value it will remain faulty even it later have the common value
  • 35. Basic Concept • If all the proper processes have the same initial message then the system will be called proper, otherwise faulty. • The processes reach agreement on the common message by exchanging information • We assume that each process Gi has a local phase- clock p(i), and that pi assigns p (i) := p (i) + 1 at the end of each phase.
  • 36. Authentication • A public directory containing for each participant B a public key' KB. • When the participant B needs to authenticate a message M, he employs a secret key DB to compute another message 0B(M) =N. • Every other user can recover M form N by use of KB, Note: The public-key directory is part of the data in each Pi, and must be incorporated by a Non-faulty "dealer" at ,the creation of the processes Gi
  • 37. LOTTERY • Algorithm employ a lottery procedure by which the proper Gi's can agree on a randomly chosen s =0 ,I. • Using Shamirs A method for obtaining digital signatures and public-key cryptosystems, • The lottery procedure admits a parameter m<N, so that Lottery(m) is the m-th lottery round. • End of the execution of Lottery(m) by the proper Gi, at least all the proper processes will share Sm.
  • 38. THE AGREEMENT PROTOCOL(BAP) Structure: Polling : Gi polls the other processes on their value of the message. Lottery: Gi decide common random value Decision: using s Gi determines whether to adopt the plurality candidate version of the message obtained through Polling or his current version of the message These 3 happen in R time. R = desired reliability
  • 42. BAP
  • 43. Correctness If Initial message of every proper process = M, Proper System: End of BAP we shall have message (i)= M for all proper Gi Faulty System: With probability at ,least 2^-R , all proper processes Gi will have the same value for message(i). Note: Probability for not reaching agreement in R rounds is 2^-R
  • 44. Bounded expected time As long as t <n/10 we can reach an agreement ignoring n and t It requires 0(n^2) message Using a wakeup call it could be reduces to 0(nt)
  • 45. Advantage • Both applied for Synchronized and asynchronous version • Better than Ben-or tossing approach • With some condition it is a robust solution for the distributed commit problem • It is more simpler than other existing approach
  • 46. Disadvantage • Synchronized system simplify the algorithm but also arises issue for coordinated agreement • For Byzanatine Consensus problem also need to implement wakeup protocol • It only resilient under appropriate conditions
  • 47. Practical Implementations • For bitcoin peer to peer this solutions could be implemented • For distributed commit problem this solution can be implemented and some cases its implemented
  • 48. Conclusion Randomize solution has some upper hand also has some limit. However in 1984 this solutions creates such effect which later used to implement various extended solutions for Byzantine agreement
  • 49. References • [1] Leslie Lamport, Robert Shostak and Marshall Pease. TheByzantine Generals Problem. ACM Transactions onProgramming Languages and Systems, 4(3): 382-401, July 1982. • [2] “AWS S3 Availability Event”, http://status.aws.amazon.com/s3-20080720.html . Amazon Rec. 10/2013 • https://en.wikipedia.org/wiki/Consensus_(computer_scie nce) • http://marknelson.us/2007/07/23/byzantine/
  • 50. References • http://lamport.azurewebsites.net/pubs/pubs.html#implem entation • http://www.cs.cornell.edu/courses/cs6410/2009fa/lecture s/24-byzantine.pdf • http://www.di.univaq.it/~proietti/slide_algdist2010/5- Consensus.ppt • http://lamport.azurewebsites.net/pubs/weak-byz.pdf • http://cs-www.bu.edu/faculty/homer/537/talks/Sachin- byzagree-lecture.pdf
  • 51. References • http://www.cs.sfu.ca/CourseCentral/401/qshi1/slides/cmp t_401_byzantine_generals.ppt • http://webcourse.cs.technion.ac.il/236357/Winter2007- 2008/ho/WCFiles/Anna- The%20Byzantine%20Generals%20Problem.ppt • http://www.cs.toronto.edu/~demke/469F.07/Lectures/Lect ure18.ppt • http://www-inst.eecs.berkeley.edu/~cs162/fa12/hand- outs/Original_Byzantine.pdf