In this Reply to the preceding Comment by Hall and Rao [Phys. Rev. A 83, 036101 (2011)], we motivate terminology of our original paper and point out that further research is needed in order to (dis)prove the claimed link between every orthogonal Latin square of order being a power of a prime and a mutually unbiased basis.
Mutually unbiased bases encapsulate the concept of complementarity---the impossibility of simultaneous knowledge of certain observables---in the formalism of quantum theory. Although this concept is at the heart of quantum mechanics, the number of these bases is unknown except for systems of dimension being a power of a prime. We develop the relation between this physical problem and the mathematical problem of finding the number of mutually orthogonal Latin squares. We derive in a simple way all known results about the unbiased bases, find their lower number, and disprove the existence of certain forms of the bases in dimensions different than power of a prime. Using the Latin squares, we construct hidden-variable models which efficiently simulate results of complementary quantum measurements.
We show there is a natural connection between Latin squares and commutative sets of monomials defining geometric structures in finite phase-space of prime power dimensions.A complete set of such monomials defines a mutually unbiased basis (MUB) and may be associated with a complete set of mutually orthogonal Latin squares (MOLS).We translate some possible operations on the monomial sets into isomorphisms of Latin squares, and find a general form of permutations that map between Latin squares corresponding to unitarily equivalent mutually unbiased sets.