Monotonicity of Fisher Discord under Partial Trace
Establish whether Fisher discord C(ρ12, H ⊗ 1) = I_F(ρ12, H ⊗ 1) − I_W(ρ12, H ⊗ 1), defined as the difference between the symmetric-logarithmic-derivative quantum Fisher information I_F and the Wigner–Yanase skew information I_W for a bipartite quantum state ρ12 and an observable H on subsystem 1, is nonincreasing under the partial trace over subsystem 2; equivalently, determine whether C(ρ12, H ⊗ 1) ≥ C(ρ1, H) holds for all bipartite states ρ12 with ρ1 = tr_2(ρ12) and all observables H on subsystem 1.
References
It is natural to consider such an open question: Whether quantum Fisher discord $$C(\rho_{12},H\otimes{\bf 1})=I_{\rm F}(\rho_{12},H\otimes{\bf 1})-I_{\rm W}(\rho_{12},H\otimes{\bf 1})$$ for bipartite states does not increase under the partial trace. If the answer is affirmative, it is worth investigating correlations of the bipartite states via the difference as $C(\rho_{12},H\otimes{\bf 1})-C(\rho_{1},H).$