Hiding in plain sight

Similar documents
Public Key Cryptography. Tim van der Horst & Kent Seamons

CS 477/677 Analysis of Algorithms Fall 2007 Dr. George Bebis Course Project Due Date: 11/29/2007

NAME: Prof. Ruiz. 1. [5 points] What is the difference between simple random sampling and stratified random sampling?

Chapter 3: Cluster Analysis

Turing Machines. Human-aware Robotics. 2017/10/17 & 19 Chapter 3.2 & 3.3 in Sipser Ø Announcement:

k-nearest Neighbor How to choose k Average of k points more reliable when: Large k: noise in attributes +o o noise in class labels

Administrativia. Assignment 1 due thursday 9/23/2004 BEFORE midnight. Midterm exam 10/07/2003 in class. CS 460, Sessions 8-9 1

Differentiation Applications 1: Related Rates

ENSC Discrete Time Systems. Project Outline. Semester

Getting Involved O. Responsibilities of a Member. People Are Depending On You. Participation Is Important. Think It Through

x 1 Outline IAML: Logistic Regression Decision Boundaries Example Data

I.S. 239 Mark Twain. Grade 7 Mathematics Spring Performance Task: Proportional Relationships

CHAPTER 24: INFERENCE IN REGRESSION. Chapter 24: Make inferences about the population from which the sample data came.

TEST 3A AP Statistics Name: Directions: Work on these sheets. A standard normal table is attached.

AP Statistics Practice Test Unit Three Exploring Relationships Between Variables. Name Period Date

AP Statistics Notes Unit Two: The Normal Distributions

READING STATECHART DIAGRAMS

Lecture 13: Markov Chain Monte Carlo. Gibbs sampling

Tree Structured Classifier

Homework #7. True False. d. Given a CFG, G, and a string w, it is decidable whether w ε L(G) True False

If (IV) is (increased, decreased, changed), then (DV) will (increase, decrease, change) because (reason based on prior research).

, which yields. where z1. and z2

Lecture 2: Supervised vs. unsupervised learning, bias-variance tradeoff

Activity Guide Loops and Random Numbers

Resampling Methods. Cross-validation, Bootstrapping. Marek Petrik 2/21/2017

CS:4420 Artificial Intelligence

CHM112 Lab Graphing with Excel Grading Rubric

Lecture 2: Supervised vs. unsupervised learning, bias-variance tradeoff

Experiment #3. Graphing with Excel

Math Foundations 20 Work Plan

Bootstrap Method > # Purpose: understand how bootstrap method works > obs=c(11.96, 5.03, 67.40, 16.07, 31.50, 7.73, 11.10, 22.38) > n=length(obs) >

Last Updated: Oct 14, 2017

Tutorial 3: Building a spectral library in Skyline

Chapter 3 Digital Transmission Fundamentals

IAML: Support Vector Machines

Lab 1 The Scientific Method

EASTERN ARIZONA COLLEGE Introduction to Statistics

SIZE BIAS IN LINE TRANSECT SAMPLING: A FIELD TEST. Mark C. Otto Statistics Research Division, Bureau of the Census Washington, D.C , U.S.A.

Hypothesis Tests for One Population Mean

Maximum A Posteriori (MAP) CS 109 Lecture 22 May 16th, 2016

SUPPLEMENTARY MATERIAL GaGa: a simple and flexible hierarchical model for microarray data analysis

T Algorithmic methods for data mining. Slide set 6: dimensionality reduction

Dataflow Analysis and Abstract Interpretation

Section 5.8 Notes Page Exponential Growth and Decay Models; Newton s Law

Fall 2013 Physics 172 Recitation 3 Momentum and Springs

AIP Logic Chapter 4 Notes

LHS Mathematics Department Honors Pre-Calculus Final Exam 2002 Answers

The Law of Total Probability, Bayes Rule, and Random Variables (Oh My!)

Physics 2010 Motion with Constant Acceleration Experiment 1

A New Evaluation Measure. J. Joiner and L. Werner. The problems of evaluation and the needed criteria of evaluation

2004 AP CHEMISTRY FREE-RESPONSE QUESTIONS

Unit 14 Thermochemistry Notes

This section is primarily focused on tools to aid us in finding roots/zeros/ -intercepts of polynomials. Essentially, our focus turns to solving.

Trigonometric Ratios Unit 5 Tentative TEST date

You need to be able to define the following terms and answer basic questions about them:

ECE 5318/6352 Antenna Engineering. Spring 2006 Dr. Stuart Long. Chapter 6. Part 7 Schelkunoff s Polynomial

1b) =.215 1c).080/.215 =.372

Revisiting the Socrates Example

Distributions, spatial statistics and a Bayesian perspective

Medium Scale Integrated (MSI) devices [Sections 2.9 and 2.10]

Department of Electrical Engineering, University of Waterloo. Introduction

**DO NOT ONLY RELY ON THIS STUDY GUIDE!!!**

Reinforcement Learning" CMPSCI 383 Nov 29, 2011!

Physics 212. Lecture 12. Today's Concept: Magnetic Force on moving charges. Physics 212 Lecture 12, Slide 1

CHAPTER 4 DIAGNOSTICS FOR INFLUENTIAL OBSERVATIONS

Sections 15.1 to 15.12, 16.1 and 16.2 of the textbook (Robbins-Miller) cover the materials required for this topic.

Modelling of Clock Behaviour. Don Percival. Applied Physics Laboratory University of Washington Seattle, Washington, USA

Exponential Functions, Growth and Decay

CONSTRUCTING STATECHART DIAGRAMS

SPECIMEN. Candidate Surname. Candidate Number

EEO 401 Digital Signal Processing Prof. Mark Fowler

Five Whys How To Do It Better

Particle Size Distributions from SANS Data Using the Maximum Entropy Method. By J. A. POTTON, G. J. DANIELL AND B. D. RAINFORD

Simple Linear Regression (single variable)

24 Multiple Eigenvectors; Latent Factor Analysis; Nearest Neighbors

Physical Layer: Outline

Please Stop Laughing at Me and Pay it Forward Final Writing Assignment

Lab #3: Pendulum Period and Proportionalities

Application Of Mealy Machine And Recurrence Relations In Cryptography

Kinetic Model Completeness

PHYS 314 HOMEWORK #3

CS 109 Lecture 23 May 18th, 2016

PHOTOSYNTHESIS THE PRACTICALS 16 APRIL 2014

Name: Block: Date: Science 10: The Great Geyser Experiment A controlled experiment

1996 Engineering Systems Design and Analysis Conference, Montpellier, France, July 1-4, 1996, Vol. 7, pp

Sequential Allocation with Minimal Switching

MODULE 1. e x + c. [You can t separate a demominator, but you can divide a single denominator into each numerator term] a + b a(a + b)+1 = a + b

Graduate AI Lecture 16: Planning 2. Teachers: Martial Hebert Ariel Procaccia (this time)

Purpose: Use this reference guide to effectively communicate the new process customers will use for creating a TWC ID. Mobile Manager Call History

THE LIFE OF AN OBJECT IT SYSTEMS

making triangle (ie same reference angle) ). This is a standard form that will allow us all to have the X= y=

Introduction to Spacetime Geometry

Functions. EXPLORE \g the Inverse of ao Exponential Function

SPH3U1 Lesson 06 Kinematics

COMP 551 Applied Machine Learning Lecture 5: Generative models for linear classification

NUMBERS, MATHEMATICS AND EQUATIONS

CHAPTER Read Chapter 17, sections 1,2,3. End of Chapter problems: 25

Chapter 8: The Binomial and Geometric Distributions

Part 3 Introduction to statistical classification techniques

Standard Title: Frequency Response and Frequency Bias Setting. Andrew Dressel Holly Hawkins Maureen Long Scott Miller

Transcription:

Hiding in plain sight Principles f stegangraphy CS349 Cryptgraphy Department f Cmputer Science Wellesley Cllege The prisners prblem Stegangraphy 1-2 1

Secret writing Lemn juice is very nearly clear s it it des nt shw n paper when dry. When heated, the lemn juice starts t burn. The released carbn shws up as brwn writing n the page. *Many thanks t Planet Science at www.scienceyear.cm. Stegangraphy 1-3 Acrstics Yur friend Wendy is nt such a bad srt. The key is t get her t tell yu abut Wellesley. She is excited abut her 15th reunin and has sme half baked idea abut ging incgnit. Wish we weren t in this place s we culd g t. I m sure that this weekend is ging t rain. Wuldn t that take the cake. Our friend waits all year fr the pprtunity t Leave wrk early and the weather at its best will be dreadful. Well, we must wait fr dawn t knw fr sure. Stegangraphy 1-4 2

Grilles The grille, credited t Gernim Cardan, imprves upn the simple acrstic. Cardan suggested cpying ut the message three times, t remve any irregularities in the size r spacing f the letters. Stegangraphy 1-5 Musical acrstics Gaspar Schtt (1608-1666) discussed hiding messages in music scres where each nte crrespnded t a letter. J.S. Bach embedded his name in the rgan chrale Vr deinen Thrn using the rule: if the ith nte f the scale ccurs k times, then the kth letter f the allphabet is t be entered in the ith place. Stegangraphy 1-6 3

Framewrk fr secret cmmunicatin Stegangraphy 1-7 Pure versus secret key stegangraphy Pure stegangraphic systems d nt require prir exchange f secret infrmatin ther than an agreement n the embedding and extractin algrithms. Secret key stegangraphy system are similar t symmetric ciphers: the sender chses a cver c and embeds the secret message int c using a secret key. Stegangraphy 1-8 4

Kerckhff s 1 st principle f cryptgraphy Assume the methd used t encipher data is knwn t the ppnent; security must lie nly in the chice f the key.* *Hw des this apply t stegangraphy? Stegangraphy 1-9 Perfect security Let P S the prbability distributin f stegcvers sent via the channel, and P C the prbability distributin f the set f cvers. A stegangraphic system is called e-secure against passive attackers, if D(P C P S ) e, and perfectly secure if e = 0. Where, P D(P C P S ) = P C (q)lg C (q) Â 2 qœq P S (q) is the relative entrpy between the tw distributins P C and P S. Stegangraphy 1-10 5

A perfectly secure steg system Therem. There exists a perfectly secure stegangraphic system. Prf. Let C be the set f all bit strings f length n, P C the unifrm distributin n C, and e ΠC the secret message. The sender selects ne c ΠC at randm and cmputes s = c XOR e. The resulting steg-cvers s are unifrmly distributed n C, s P C = P S. Stegangraphy 1-11 Detecting secret messages T decide whether a cver cntains secret infrmatin r nt Wendy defines a test functin secretinf: C -> {true, false}. Wendy makes a type-ii errr (prbability b) if she fails t detect a hidden message. She makes a type-i errr (prbability a) if she falsely detects a hidden message in a cver that has nne. Stegangraphy 1-12 6

Cachin s Therem Therem. In an e-secure steg system, a lg 2 a 1 - b + (1 -a)lg 1 -a 2 e b In particular, if a = 0, then b 2 -e Stegangraphy 1-13 Hiding in the nise The general principle underlying mst stegangraphic methds is t place the secret message in the nise cmpnent f a signal. If this is pssible in such a way that the result is indistinguishable frm true randm nise, an attacker has n chance f detecting a secret cmmunicatin. Stegangraphy 1-14 7

Wendy ges n the ffensive Stegangraphic systems are extremely sensitive t cver mdificatins. An active attacker, wh is nt able t extract r prve the existence f a secret message might simply add randm nise t the transmitted cver and s try t destry the infrmatin. Stegangraphy 1-15 Rbustness A secret key stegangraphic system is P-rbust fr sme class P f mappings C -> C, if fr all p ΠP D K (p(e K (c, m))) = D K (E K (c,m)) = m. There is a trade-ff between security and rbustness. The mre rbust a system, the less secure since rbustness can nly be achieved by redundant infrmatin. Stegangraphy 1-16 8

Resistance t minr changes Supraliminal channel: Infrmatin hidden in plain sight, s bviusly, in fact that it is impssible t mdify withut grss mdificatin f the transmitted bject. Cver-plt: A frmal descriptin f the perceptually significant parts f the cver. Cver-plt functin: A mapping f frm cver-plts t {0, 1} n. Stegangraphy 1-17 Alice uses a supraliminal channel Stegangraphy 1-18 9