Technical Summary: Entanglement Summoning from Entanglement Sharing

Problem Statement

This work addresses the entanglement summoning problem, a scenario within the quantum tasks framework where distributed, cooperating parties must prepare a maximally entangled state between two specific, distant locations in response to time-sensitive requests. The parties operate under strict causal constraints defined by a directed graph (the causal graph), which dictates the allowed communication links and timing.

The core challenge is to determine the necessary and sufficient conditions on the causal graph that allow a set of parties to successfully complete the task. Specifically, given a causal graph $G_C$ and a set of "diamonds" (spacetime regions defined by causal pasts and futures), the task requires that if exactly two diamonds receive a "call" (input bit 1), the parties can output a maximally entangled state between those two locations using only local operations and the allowed communication channels.

Previous work [1, 2] fully characterized the case where all causal connections are oriented (one-way, i.e., $A \to B$ or $B \to A$, but not both). This paper extends the characterization to:

1. The case where all connections are bidirected ($A \leftrightarrow B$).
2. The general mixed case containing both oriented and bidirected edges, for which the authors provide sufficient conditions but leave the necessity of some conditions open.

Methodology

The authors' primary strategy is to reduce the dynamical problem of entanglement summoning to the static problem of entanglement sharing.

1. Reduction to Entanglement Sharing Schemes (ESS):
   Building on the development of entanglement sharing schemes [3], the authors construct a protocol where the existence of a successful summoning strategy is equivalent to the existence of a specific quantum state (an ESS) that satisfies certain access structures. In an ESS, non-communicating parties hold subsystems, and specific pairs of subsets must be able to recover a maximally entangled state via local operations.

2. Graph Theoretic Mapping:
   The authors map the causal graph $G_C$ of the summoning task to an access-pair graph $G_A$ (an undirected graph) associated with the ESS.

   • Bidirected Case: For a purely bidirected causal graph, the access-pair graph is simply the complement of the causal graph (treating bidirected edges as undirected).
   • General Case: For mixed graphs, the authors define a more complex construction (Algorithm 34) that introduces "body vertices" (corresponding to diamonds) and "wing vertices" (corresponding to specific causal dependencies) to form the access-pair graph.

3. Application of Monogamy Conditions:
   The solvability of the ESS is governed by the monogamy condition (Theorem 10 [3]), which relates the graph structure to the overlap of subsystems. Specifically, for any even-length path in the access-pair graph, the start and end subsets must overlap. The authors translate this graph-theoretic condition back into constraints on the original causal graph.

Key Contributions and Results

1. Characterization of Bidirected Entanglement Summoning

The paper provides a complete "if and only if" characterization for summoning tasks where all causal edges are bidirected.

• Theorem 1 & Theorem 30: An entanglement summoning task with a bidirected causal graph is achievable if and only if the causal graph admits a two-clique partition.
• Graph Theoretic Equivalence: This condition is equivalent to stating that the complement of the causal graph contains no odd cycles (i.e., the complement is bipartite).
• Interpretation: The two cliques in the partition correspond to the two subsystems of the maximally entangled state. The protocol involves moving one half of the entangled pair through the first clique and the other half through the second clique, utilizing the fact that a clique contains a tournament as a spanning subgraph (allowing single-system summoning).

2. Sufficient Conditions for General (Mixed) Summoning Tasks

For the general case involving both oriented and bidirected edges, the authors provide a set of sufficient conditions (Theorem 43 and Corollary 44) based on the structure of the causal graph.

The conditions are:

• (NOC*) The causal graph admits a two-quasi-clique partition (a partition into two sets where every pair of vertices within a set has at least one directed edge between them).
• (M1**) If $D_i \not\leftarrow D_j$ (no causal connection from $i$ to $j$), there exists a two-quasi-clique partition where $D_i$ and $D_j$ belong to different cliques.
• (M3**) If $D_{i1}$ and $D_{i2}$ belong to the same quasi-clique in every two-quasi-clique partition, and $D_{i1} \not\leftarrow D_{j1}$ and $D_{i2} \not\leftarrow D_{j2}$, then $D_{j1}$ and $D_{j2}$ must be connected ($D_{j1} \sim D_{j2}$).

These conditions are derived by translating the monogamy constraints of the associated access-pair graph back into the language of the causal graph.

3. Necessity and Open Questions

• Proven Necessity: The authors prove that (NOC*) and (M2*) (a condition related to the "no two-out" scenario) are necessary for any achievable task.
• Open Necessity: The necessity of conditions (M1*) and (M3*) in the general mixed case remains an open question. The authors note that their protocol, which maps the problem to an ESS, may not be fully general for mixed graphs.
• Uncharacterized Example: The paper highlights Figure 2 as the simplest uncharacterized entanglement summoning problem. In this example, removing a single directed edge makes the task impossible, while making it bidirected makes it solvable, illustrating the complexity of the mixed case.

Significance and Scope

The paper claims to advance the understanding of the limits of entanglement structure in many-body quantum states under relativistic constraints.

• Theoretical Framework: It extends the "quantum tasks" framework by providing a rigorous characterization of entanglement summoning in bidirected networks, a significant step beyond the previously solved oriented case.
• Connection to Entanglement Sharing: The work solidifies the link between dynamic summoning tasks and static entanglement sharing schemes, demonstrating that the former can often be reduced to the latter.
• Modest Claims: The authors explicitly state that while they provide a complete characterization for the bidirected case, their results for the general mixed case are sufficient but not proven necessary. They do not claim to have solved the general case completely, leaving the necessity of the derived conditions and the existence of alternative protocols for mixed graphs as open problems for future research.

The work relies on the foundational results of Kent et al. [1, 2] regarding oriented graphs and Khanian et al. [3] regarding entanglement sharing schemes, integrating them to address the specific constraints of bidirected and mixed causal networks.