This paper presents a infeasible interior-point primal -dual affine scaling algorithm for linear programming. it is shown that the method is polynomial-time algorithm.

Some numerical results for a large number of random convex quadratic programming problems show that the new algorithm is efficient and might be a polynomial-time algorithm under some conditions.

In this paper, we present a new match protocol. It is of semantic security: there exists no probabilistic polynomial time algorithm to distinguish two guessed inputs.