Properties of computational entanglement measures

Abstract

Quantum entanglement is a useful resource for implementing communication tasks. However, for the resource to be useful in practice, it needs to be accessible by parties with bounded computational resources. Computational entanglement measures quantify the usefulness of entanglement in the presence of limited computational resources. In this paper, we analyze systematically some basic properties of two recently introduced computational entanglement measures, the computational distillable entanglement and entanglement cost. To do so, we introduce lower bound and upper bound extensions of basic properties to address the case when entanglement measures are not defined by a scalar value but when only lower or upper function bounds are available. In particular, we investigate the lower bound convexity and upper bound concavity properties of such measures, and the upper and lower bound additivity with respect to the tensor product. We also observe that these measures are not invariant with local unitaries, although invariance is recovered for efficient unitaries. As a consequence, we obtain that these measures are only LOCC monotones under efficient families of LOCC channels. Our analysis covers both the one-shot scenario and the uniform setting, with properties established for the former naturally extending to the latter.