1. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)1 Encryption A Brief Overview

2. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)2 Acknowledgements Charles Pfleedger E. Spafford William A. Stein FOLDOC Sunit Chauhan Jim Xu, et al. Shawn Hillis

3. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)3 Outline Basic concepts Stream ciphers Block ciphers Summary of Stream and Block ciphers Public key encryption

4. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)4 Why do we need encryption Scenario: –S wants to send the message T to R, where an outsider, O, wants the message and tries to access it. –S: Sender –R: Receiver –T: Transmission Medium –O: Interceptor or Intruder. 4 ways O might try to access message. –Block it: prevent T from reaching R (availability) –Intercept it: read or listen to message (secrecy) –Modify it: obtaining message and changing it –Fabricate: generate an authentic-looking message to be delivered to R appearing to come from S

5. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)5 Basic Concepts Encryption: –Definition: mechanisms to disguise the message so that if the intermission is intercepted/diverted, the content of the message will not be understood. –Impact: foundational building block to security-based computing

6. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)6 Basic Concepts – cont’d Cryptosystem –5-tuple M: set of plain text C: set of cipher text K: set of keys E: M x K -> C D: C x K -> M

7. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)7 Basic Concepts – cont’d Encryption: process of encoding a message so that its meaning is not obvious Decryption: transforming encrypted message back to its normal form Encode/decode: translating phrases to other words or phrases Encipher/decipher: translating letters or symbols individually. Plaintext: original form of message: P = (p1,p2,…, pn) Ciphertext: encrypted form of message: C = (c1,c2,…, cn) Encryption/decryption relationships: –C = E(P); P = D(C); P = D(E(P))

8. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)8 Basic Concepts – cont’d Some encryption algs use a key K –C = E(K,P) –E is a SET of encryption algs –Key K selects specific one Symmetric Encryption: P = D(K,E(K,P)) –encryption/decryption keys are the same Asymmetric Encryption: P = D(K D,E(K E,P))

9. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)9 Pictorial Representation Key PlaintextCiphertext Original Plaintext EncryptionDecryption Encryption Key PlaintextCiphertext Original Plaintext EncryptionDecryption Decryption Key KDKD KEKE Symmetric Encryption: Asymmetric Encryption:

10. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)10 Basic Concepts – cont’d Cryptography: (hidden writing) –Practice of using encryption to conceal text Cryptanalyst: –Person who studies encryption and encrypted messages –Intent: find hidden meaning Cryptographer and Cryptanalyst: –Both attempt to translate coded material to original form –Cryptographer: works on behalf of legitimate sender or receiver. –Cryptanalyst: Works on behalf of unauthorized interceptor Cryptology: research/study into encryption/decryption –Includes cryptography and cryptanalysis.

11. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)11 Basic Concepts – cont’d Cryptanalysis –Objective: Break an encryption Deduce the meaning of a ciphertext mesg Determine decrypting algorithm that matches an encrypting algorithm –Possible techniques: break single message Recognize patterns in encrypted mesgs –break subsequent mesgs with straightforward decryption alg Find general weaknesses in encryption alg –Without necessarily intercepting any mesgs –Tools: Encrypted mesgs, known encryption algs, intercepted plaintext, data elements known/suspected of being in ciphertext, mathematical/statistical techniques, props of languages, computers, and luck

12. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)12 Basic Concepts – cont’d Breakability –Encryption algorithm is BREAKABLE: Given enough time and data, an analyst could determine the algorithm Practicality is issue For given cipher scheme, may have 10 30 possible decipherments –Select one from 10 30 Current technology: perform 10 10 ops/sec –Require 10 20 secs – 10 12 years –Reality Check: Cryptanalyst won’t just try the “hard” ways –Ex: more clever approach, might only take 10 15 ops 10 10 ops/sec, 10 15 ops will take about one day Breakability estimates are based on CURRENT technology

13. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)13 Stream Ciphers and Block Ciphers Stream Ciphers –Letter by letter –E.g. substitution-based cipher Block cipher –Block by block –E.g. transposition based cipher

14. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)14 Character Representations Study ways to encrypt any computer material: –ASCII/EBCDIC chars –Binary data or Object code –Control stream ABCDEFGHIJKLM 0123456789101112 NOPQRSTUVWXYZ 13141516171819202122232425

15. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)15 Substitution-based Encryption Monoalphabetic Ciphers –Caesar Cipher: c i = E(p i ) = p i + k –Cryptosystem K: {k| 0<=k<=25} E: (m+k) mod 26 D: (26+c-k) mod 26 C = M = {all sequence of roman letters} –Examples ( k=3) wuhdwb lpsrvvleoh, wklv phvvdjh lv qrw wrr kdug wr euhdn –Evaluation Easy to perform in field (no written instructions) Too simple, very easy to break

16. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)16 Substitution-based Encryption – cont’d –Weakness: study frequency distribution [Jim Xu, et al.]

17. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)17 Substitution-based Encryption-cont’d Polyalphabetic Substitution Ciphers –Desire flat distribution –Combine distributions that are high with low ones Encipher T as a and sometimes as b Also encipher X as a and sometimes as b –Use two separate encryption alphabets Tables for odd and even positions    mod 26    mod 26 TREAT YIMPO SSIBL E Fumnf dyvtf czysh h –Weaknesses Break by studying patterns (digram, trigram)

18. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)18 Substitution Discussion Major weakness: –frequency distribution (index of coincidence: measure of variation between frequencies in a distribution) –Some letters are just used more frequently than others –Numerous enciphering techniques still can make it difficult to hide these patterns –Kasiski Method: find number of alphabets used Identify repeated patterns of 3 or more chars For each pattern, write down position at which each instance of pattern begins Compute difference between start points of success instances Determine all factors of each difference If polyalphabetic subst used, key length will be one of the factors that appears often in previous step.

19. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)19 Transpositions (Permutations) Definition: encryption where letters are rearranged. Goal: diffusion, spread info from message or key out widely across the ciphertext. Try to break established patterns. –Digram or trigram: patterns of adjacent letters. Study 2 and 3 letter combinations of adj letters E.g. th, en, ing, etc.

20. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)20 Transposition Techniques Columnar Transpositions: –Rearrangement of chars of plaintext into cols C1C2C3C4C5 C6C7C8C9C10 C11C12Etc. THISI SAMES SAGET OSHOW HOWAC OLUMN ARTRA NSPOS ITION WORKS tssoh oaniw haaso lrsto imghw utpir seeoa mrook istwc nasns

21. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)21 Transposition Double Transposition Alg: –Involves 2 columnar transpositions –With different number of columns, applied sequentially. Fractionated Morse: –keyed mono-alphabetic cipher –Result is subsequently blocked (clustered) –Morse code is used as its basis

22. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)22 Summary Stream Ciphers (substitution based) –Mono and poly-alphabetic encryptions –Since they convert one symbol of plaintext immediately into a symbol of ciphertext –Features include: a) speed of transformation, b) low error propagation –Disadvantages are: a) low diffusion, b) susceptible to malicious insertions

23. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)23 Summary – cont’d Block Ciphers –Columnar transposition algorithm and fractionated Morse –Since a group of plaintext symbols are encrypted as one block –Features include: a) diffusion, b) immunity to insertions –Disadvantages are: a) slowness of encryption, b) error propagation

24. R R R CSE870: Advanced Software Engineering: Cheng (Sp 2003)24 Summary – cont’d Combination of stream cipher and block cipher can be very powerful –DES –Will go into details