Computational vs. Perfect Secrecy

Computational - what we're studying!!!

all adversaries can Brute Force Search. GOALS:

• Key space Exponential in length
  • make it infeasable to brute force search (key space too big) - use something >2^100
• Brute Force = worst case. make it the best strategy for the adversary
  • all keys and messages are equally likely

• Short fixed key length
• Crackable under computing power
• too many potential keys

Perfect Secrecy (information theoretic)

• Long, message length key
• CAN'T be cracked, even with finite power since all messages are equally possible
• key transfer and management totally impractical for most real world applications