Art of the Problem | Gambling with Secrets: Part 6/8 (Perfect Secrecy & Pseudorandomness) @ArtOfTheProblem | Uploaded 12 years ago | Updated 11 hours ago
This chapter introduces why random shifts result in perfect secrecy. We explore hardware random number generators vs. pseudorandom number generators which expand a short random seed into a long sequences of "random looking" numbers. If used in cryptography, can offer "practical security" which is based on the secrecy of the random seed only. Alice and Bob can safely assume that the "enemy knows the system" (one-time pad + pseudorandom generator), and focus on assuring their seed is shared in secret & generated randomly.
If you are still wondering: how exactly would Eve break a ciphertext encrypted with a pseudorandom list of shifts? Remember: the pseudorandom generator can only produce a tiny (0.0000...1%) fraction of all possible shift sequences. Eve could program a computer to generate all possible sequences (starting from seed 0, 1, 2, 3.....) and reverse all of these shifts. When the computer hits a seed which decrypts the ciphertext into an "English looking" sentence, she can be 99.999% sure it is the message. All of the other seeds will decrypt into unreadable (jumbled looking) messages.
This chapter introduces why random shifts result in perfect secrecy. We explore hardware random number generators vs. pseudorandom number generators which expand a short random seed into a long sequences of "random looking" numbers. If used in cryptography, can offer "practical security" which is based on the secrecy of the random seed only. Alice and Bob can safely assume that the "enemy knows the system" (one-time pad + pseudorandom generator), and focus on assuring their seed is shared in secret & generated randomly.
If you are still wondering: how exactly would Eve break a ciphertext encrypted with a pseudorandom list of shifts? Remember: the pseudorandom generator can only produce a tiny (0.0000...1%) fraction of all possible shift sequences. Eve could program a computer to generate all possible sequences (starting from seed 0, 1, 2, 3.....) and reverse all of these shifts. When the computer hits a seed which decrypts the ciphertext into an "English looking" sentence, she can be 99.999% sure it is the message. All of the other seeds will decrypt into unreadable (jumbled looking) messages.
