Perfect random number generator is unnecessary for secure quantum key distribution

Xiang-Bin Wang
2004-01-01
Abstract:Game G: Clare passes a string s which is either from perfect random number generator R0 or from good imperfect number generator R1, with equal probability. Alice's information about whether it is from R0 or R1 is bounded by small value h. Alice use s as the input random numbers for QKD protocol with Bob. Suppose Eve may have very small information about the final key if s is from R0 and Eve has large information if s is from R1, then after the protocol, Alice announce the final key, Eve's information about whether s is from R0 or R1 is unreasonablly large, i.e., breaks the known bound, h. Explicit formulas are given in the article.
What problem does this paper attempt to address?