Cryptography Project
Home
ABSTRACT

One of the fundamentals of cryptography is that keys or secret numbers are
selected and used which are computationally infeasible for an attacker to
compute the same key given the public information. Consider one of the most
commonly used assumptions for cryptography, the RSA assumption, which states
given a large number n = p*q such that p and q are primes, e such that
GCD(e, ?(n)) = 1, and ciphertext C, it is computationally infeasible
to compute the original message M such that C = M^e mod N. There
are many other assumptions made by cryptographic protocols which rely on the
computational infeasibility of an attacker having the ability to produce
secret keys of different sorts. However, if there was a way to drastically
increase the computational power of a machine these assumptions would not
necessarily hold true. Many people optimistically speculate that one way to
achieve this increase in computing power is by looking into quantum
computing. What effects will this have on cryptography? Will we be able to
enhance the current strategies of cryptographic methods or will new
strategies have to be created? These are questions which, rightfully so,
are already being asked as we verge on the barrier of the quantum
computing capabilities.


Paper: The Effects of Quantum Computers on Cryptography (pdf)

Paper: Quantum Computers and Cryptography (ppt)