In a quantum network whose edges share partially entangled two-qubit pure states, entanglement percolation asks when local operations and classical communication (LOCC) can establish a maximally entangled pair between two arbitrary distant nodes. Romancino, Romero-Pallejà, Palma and Sanpera (Quantum Sci. Technol. 11, 025044, 2026) extended this framework to heterogeneous networks in which each edge's singlet-conversion probability (SCP) is drawn at random from a distribution. They proved that the random classical entanglement percolation (RCEP) threshold depends only on the mean SCP, while the random quantum entanglement percolation (RQEP) threshold degrades linearly with the distribution width, yielding a crossover width w∗w^* beyond which RCEP outperforms RQEP, and conjectured that RCEP is the optimal LOCC protocol when the variance of the SCP distribution is sufficiently large. The open problem is to characterize the optimal LOCC strategy and the fundamental entanglement thresholds for perfect long-distance entanglement in heterogeneous quantum networks — prove or refute the RCEP-optimality conjecture, establish general conditions and topology-dependent expressions for the crossover width w∗w^*, and determine whether a universal minimum initial entanglement exists below which no LOCC protocol can establish perfect long-distance entanglement.

Background

A quantum network is modeled as a graph whose vertices are nodes capable of local quantum operations and whose edges are shared quantum states, here partially entangled two-qubit pure states ∣ω⟩=λ1∣00⟩+λ2∣11⟩|\omega\rangle=\sqrt{\lambda_1}|00\rangle+\sqrt{\lambda_2}|11\rangle with Schmidt coefficients λ1+λ2=1\lambda_1+\lambda_2=1 (the Schmidt decomposition of a bipartite pure state). The goal of entanglement percolation is to establish a maximally entangled singlet between two arbitrary distant nodes using only local operations and classical communication (LOCC), i.e. local measurements coordinated by classical messages. By Nielsen's majorization theorem, a pure two-qubit state can be converted to a singlet by LOCC with probability pω=2λ2p_\omega=2\lambda_2, called the singlet conversion probability (SCP). In classical entanglement percolation (CEP), each edge is converted to a singlet with its SCP, mapping the network onto bond percolation, the classical statistical-mechanics model in which edges of a lattice are independently open with probability pp and a percolation threshold pcp_c separates a phase with a spanning connected cluster from a phase without one; two nodes can share a singlet whenever they lie in the same cluster, because entanglement swapping (a joint measurement at an intermediate node that transfers entanglement to its two neighbors) propagates singlets along a path. Quantum entanglement percolation (QEP) improves on this by preprocessing the lattice with local quantum operations — most prominently qq-swaps, entanglement-swapping operations performed at nodes that join groups of incident edges and transform the lattice into a different one with a lower percolation threshold (Acín, Cirac and Lewenstein, Nature Physics 3, 256, 2007). However, the optimal protocol for a given network has remained an open question since that founding work; optimal LOCC measurements are known only for small one- and two-dimensional networks and hierarchical lattices (Perseguers et al., Phys. Rev. A 77, 022308, 2008), and related studies cover complex-network topologies (Cuquet and Calsamiglia, Phys. Rev. Lett. 103, 240503, 2009; Phys. Rev. A 83, 032319, 2010), multipartite measurements (Perseguers et al., Phys. Rev. A 81, 032327, 2010), restricted LOCC classes (Khanna, Halder and Sen, Phys. Rev. A 109, 012419, 2024), small-world networks (Liang, Chen and Wang, arXiv:2401.11862, 2024), and heuristic swapping-plus-distillation algorithms for planar networks with quenched disorder (De Girolamo, Magnifico and Lupo, Quantum Sci. Technol. 10, 035047, 2025). Realistic networks are heterogeneous: link lengths, losses and noise differ, so edge entanglement is not identical. Romancino, Romero-Pallejà, Palma and Sanpera (Quantum Sci. Technol. 11, 025044, 2026) introduced random entanglement percolation, where each edge's SCP is drawn independently from a probability distribution of mean ⟨p⟩\langle p\rangle and width ww. They proved that the RCEP threshold equals the CEP threshold of the mean, pRCEP=pCEP(⟨p⟩)p^{\mathrm{RCEP}}=p^{\mathrm{CEP}}(\langle p\rangle) — only the average initial entanglement matters — while the RQEP threshold, computed via the distribution of the minimum SCP entering a qq-swap, degrades linearly with the width for location-scale families of SCP distributions; for a double-bond honeycomb lattice mapped to a triangular one, pRQEP=pQEP(⟨p⟩)+w/6≈0.347+w/6p^{\mathrm{RQEP}}=p^{\mathrm{QEP}}(\langle p\rangle)+w/6\approx 0.347+w/6, so a crossover width w∗≈0.067w^*\approx 0.067 exists beyond which pRCEP<pRQEPp^{\mathrm{RCEP}}<p^{\mathrm{RQEP}}. On this basis they conjectured that RCEP is the optimal LOCC protocol for entanglement distribution when the variance of the SCP distribution is sufficiently large, and left open whether a minimum amount of initial entanglement is needed at all to achieve perfect long-distance entanglement.

Problem Statement

For a quantum network whose edges independently share partially entangled two-qubit pure states with singlet-conversion probabilities (SCPs) drawn from a distribution of mean ⟨p⟩\langle p\rangle and width ww (a random-SCP or heterogeneous quantum network), determine the optimal LOCC (local operations and classical communication) strategy for entanglement percolation and the entanglement thresholds it induces. Following the source paper (Romancino et al., 2026), the width is w=b−aw=b-a for a uniform U(a,b)U(a,b) SCP distribution and, more generally, the scale parameter of a location-scale family of distributions. Concretely, with pRCEP=pCEP(⟨p⟩)p^{\mathrm{RCEP}}=p^{\mathrm{CEP}}(\langle p\rangle) denoting the random classical entanglement percolation threshold and pRQEPp^{\mathrm{RQEP}} the threshold under qq-swap-based random quantum entanglement percolation: (1) prove or refute the conjecture of Romancino et al. (2026) that RCEP is the optimal LOCC protocol for entanglement distribution when the variance of the SCP distribution is sufficiently large — in particular, either exhibit an LOCC protocol (possibly beyond qq-swaps) that outperforms RCEP in the high-variance regime on some network, or prove that none exists; (2) establish, beyond the single worked example of the double-bond honeycomb lattice with a uniform SCP distribution (where pRQEP≈0.347+w/6p^{\mathrm{RQEP}}\approx 0.347+w/6 and w∗≈0.067w^*\approx 0.067), general conditions under which a crossover width w∗w^* separating the QEP-favorable and CEP-favorable regimes exists, together with tight bounds or topology-dependent expressions for it; and (3) determine whether there exists a minimum amount of initial entanglement below which no LOCC protocol whatsoever can establish perfect long-distance entanglement on such networks. These three questions jointly serve one objective: a theory of LOCC-optimal entanglement distribution under heterogeneous quantum-network resources.

References

  1. A. Acín, J. I. Cirac, M. Lewenstein, "Entanglement percolation in quantum networks", Nature Physics 3, 256–259 (2007). DOI: 10.1038/nphys549, https://doi.org/10.1038/nphys549
  2. S. Perseguers, J. I. Cirac, A. Acín, M. Lewenstein, J. Wehr, "Entanglement distribution in pure-state quantum networks", Physical Review A 77, 022308 (2008). DOI: 10.1103/PhysRevA.77.022308, arXiv:0708.1025, https://arxiv.org/abs/0708.1025
  3. M. Cuquet, J. Calsamiglia, "Entanglement percolation in quantum complex networks", Physical Review Letters 103, 240503 (2009). DOI: 10.1103/PhysRevLett.103.240503, arXiv:0906.2977, https://arxiv.org/abs/0906.2977
  4. S. Broadfoot, U. Dorner, D. Jaksch, "Entanglement Percolation with Bipartite Mixed States", EPL 88, 50002 (2009). DOI: 10.1209/0295-5075/88/50002, arXiv:0906.1622, https://arxiv.org/abs/0906.1622
  5. M. Cuquet, J. Calsamiglia, "Limited-path-length entanglement percolation in quantum complex networks", Physical Review A 83, 032319 (2010). DOI: 10.1103/PhysRevA.83.032319, arXiv:1011.5630, https://arxiv.org/abs/1011.5630
  6. S. Perseguers, D. Cavalcanti, G. J. Lapeyre, M. Lewenstein, A. Acín, "Multipartite entanglement percolation", Physical Review A 81, 032327 (2010). DOI: 10.1103/PhysRevA.81.032327, arXiv:0910.2438, https://arxiv.org/abs/0910.2438
  7. S. Perseguers, G. J. Lapeyre, D. Cavalcanti, M. Lewenstein, A. Acín, "Distribution of entanglement in large-scale quantum networks", Reports on Progress in Physics 76, 096001 (2013). DOI: 10.1088/0034-4885/76/9/096001, https://doi.org/10.1088/0034-4885/76/9/096001
  8. X. Meng, J. Gao, S. Havlin, "Concurrence Percolation in Quantum Networks", Physical Review Letters 126, 170501 (2021). DOI: 10.1103/PhysRevLett.126.170501, https://doi.org/10.1103/PhysRevLett.126.170501
  9. O. Malik, X. Meng, S. Havlin, G. Korniss, B. K. Szymanski, J. Gao, "Concurrence percolation threshold of large-scale quantum networks", Communications Physics 5, 193 (2022). DOI: 10.1038/s42005-022-00958-4, https://doi.org/10.1038/s42005-022-00958-4
  10. G. Bianconi, S. N. Dorogovtsev, "The theory of percolation on hypergraphs", arXiv:2305.12297 (2023). arXiv:2305.12297, https://arxiv.org/abs/2305.12297
  11. S. Khanna, S. Halder, U. Sen, "Quantum entanglement percolation under a realistic restriction", Physical Review A 109, 012419 (2024). DOI: 10.1103/PhysRevA.109.012419, arXiv:2008.09040, https://arxiv.org/abs/2008.09040
  12. J. Liang, X. Chen, Y. Wang, "Effects of Different Q-swaps Modes on Percolation Threshold in Small-world Quantum Networks", arXiv:2401.11862 (2024). arXiv:2401.11862, https://arxiv.org/abs/2401.11862
  13. D. Nath, S. Roy, "General concurrence percolation on quantum networks", Physical Review A 114, 012440 (2026). DOI: 10.1103/2sf7-srwz, arXiv:2501.11004, https://arxiv.org/abs/2501.11004
  14. A. De Girolamo, G. Magnifico, C. Lupo, "Percolation thresholds and connectivity in quantum networks", Quantum Science and Technology 10, 035047 (2025). DOI: 10.1088/2058-9565/ade55e, arXiv:2502.17191, https://arxiv.org/abs/2502.17191
  15. Y. Zhao, K. He, Y. Zhang, J. Hou, J. Gao, S. Havlin, X. Meng, "Negativity Percolation in Continuous-Variable Quantum Networks", arXiv:2507.16417 (2025). arXiv:2507.16417, https://arxiv.org/abs/2507.16417
  16. L. Cirigliano, V. Brosco, C. Castellano, S. Felicetti, L. Pilozzi, B. van Heck, "Dynamical entanglement percolation with spatially correlated disorder", arXiv:2601.05925 (2026). arXiv:2601.05925, https://arxiv.org/abs/2601.05925
  17. A. Romancino, J. Romero-Pallejà, G. M. Palma, A. Sanpera, "Entanglement percolation in random quantum networks", Quantum Science and Technology 11, 025044 (2026). DOI: 10.1088/2058-9565/ae5fca, arXiv:2602.10189, https://arxiv.org/abs/2602.10189
  18. A. Romancino, "Random entanglement percolation on realistic quantum networks", Il Nuovo Cimento C 49, 176 (2026). DOI: 10.1393/ncc/i2026-26176-2, arXiv:2604.21967, https://arxiv.org/abs/2604.21967