Information Hiding & Digital Watermarking Tri Van Le
Outlines Some history State of the art Research goals Possible approaches Research plan
Cryptography in the 80s Beginning time of open research A lot of schemes proposed Most of them soon broken
Broken Cryptosystems (I) Merkle Hellman Iterated Knapsack Lu-Lee Merlke Hellman Merlke Hellman Lu-Lee Adiga Shankar Adigar Shankar Nieder- reiter Neiderreiter Goodman McAuly Goodman McAuly Pieprzyk Pieprzyk Chor Rivest Chor Rivest Okamoto Okamoto Okamoto
Broken Cryptosystems (II) Matsumoto Imai Cade Yagisawa Matsumoto Imai CadeYasigawa TMKIF Tsujii, Itoh Matsumoto Kurosama Fujioka Luccio Mazzone Luccio Mazzone Kravitz Reed Kravitz Reed Rao Nam Rao Nam Low Degree CG 1982 High Degree CG 1988 Rivest Adleman Dertouzos Rivest Adleman Dertouzos Krawczyk Boyar
Broken Cryptosystems (III) Ong Schnorr Ong Schorr Ong Schnorr Shamir Ong Schorr Shamir Okamoto Shiraishi Okamoto Shiraishi
Proven Secure Cryptosystems (I) Shannon’s work (1949) –Mathematical proof of security –Information theoretic secrecy Enemy with unlimited power –Can compute any desired function
Proven Secure Cryptosystems (II) Rabin (81), Goldwasser & Micali (82) –Mathematical proof of security –Computational secrecy Enemy with limited time and space –Can run in polynomial time –Can use polynomial space
Information Hiding (state of the art) Similar to that of cryptography in 80s –Many schemes were proposed –Most of them were broken Use heuristic security –Subjective measurements –Assume very specific enemy
Broken Schemes (I)
Broken Schemes (II)
Broken Schemes (III)
Broken Schemes (IV)
Research Goals Fundamental way –Systematic research –Same as Shannon and Goldwasser’s work What have been done –Covert channels –Anonymous communications What are the properties
Fundamental Models Unconditional hiding –Unlimited enemy Statistical hiding –Polynomial samples Computational hiding –Polynomial time
What have been done Covert channels Anonymous communications Information hiding –Steganography –Digital watermarking
Covert Channels Leakage information (e.g. viruses) –Disk space –CPU load Subliminal channels –Digital signatures –Encryption schemes –Cryptographic malwares
Covert Computations Computation inside computations –Secret design calculations inside a factoring computation –Secret physics simulations inside a cryptographic software or devices
Anonymous Communications MIX Networks –Electronic voting –Anonymous communication Onion Routings –Limited anonymous communication Blind signatures –Digital cash
Information Hiding Steganography –Invisible inks –Small dots –Letters Digital watermarking –Common lossy compressions –Common signal processing operations
Information Hiding Hiding property –Output must look like the cover Secrecy –No partial information on input message Authenticity –Hard to compute valid output
Our Approaches Arbitrary key –Steganography, watermarking Restricted key –Protection of key materials Key = Ciphertext –Secret sharing
Research Plan To understand information hiding –Perfect hiding (done) Necessary and sufficient conditions Computational complexity results Constructions of prefect secure schemes Constructions of schemes with non-reliability –Computational hiding (under research) Conventional constructions Public key schemes
Research Plan Other aspects –Replacing privacy by authenticity Extra problem –Robustness against modifications
Thank you Questions? More details?
Approaches Arbitrary key distribution –E: K M C –K: key space –M: message space –C: cover space Requires –E(k,m) is distributed accordingly to P cover
Approaches Restricted key distribution –c = E(k,m) –k is distributed accordingly to P K –c is distributed accordingly to P Cover
Approaches Key = Ciphertext –S: M C C –(k 1,k 2 ) = S(m) Requires –k 1 and k 2 distributed accordingly to P Cover
Models Perfect hiding –P c = P cover –Ciphertext distributes exactly as P cover Statistical Hiding –|P c - P cover | is a negligible function Negligible function –f(n) 0 and n>N d.
Models Computational Hiding –P c and P cover are P-time indistinguishable –For all P-time P.T.M. M: Prob(M(P c )=1) - Prob(M(P cover )=1) is negligible.
Examples Quadratic residues –n = pq –S 1 = {x 2 |x in Z n * } –S 2 = {x|x in Z n * and J(x)=1} Decision Diffie-Hellman –U 1 = (g, g a, g b, g ab ) mod p –U 2 = (g, g a, g b, g r ) mod p