How many copies are needed for state discrimination?
arXiv:quant-ph/0606131 · doi:10.1109/TIT.2011.2169544
Abstract
Given a collection of states (rho_1, ..., rho_N) with pairwise fidelities F(rho_i, rho_j) <= F < 1, we show the existence of a POVM that, given rho_i^{otimes n}, will identify i with probability >= 1-epsilon, as long as n>=2(log N/eps)/log (1/F). This improves on previous results which were either dimension-dependent or required that i be drawn from a known distribution.
1 page, submitted to QCMC'06, answer is O(log # of states)