Unconstrained distillation capacities of a pure-loss bosonic broadcast channel

I. INTRODUCTION

• Loss severely limits the rates and distances at which secret key or entanglement can be distilled, with standard optical-fiber secret-key rates dropping exponentially with distance.The observed rate-loss tradeoff is a fundamental limit even with unconstrained input energy.Source: §I · p. 1
• Refs., [10] proved an upper bound on the two-way LOCC-assisted quantum and secret-key agreement capacity of a pure-loss bosonic channel as a function of transmittance η; Ref. [13] established unconstrained capacity C(η) = −log2(1−η).The constrained capacity when transmitter energy is finite remains open.Source: §I · p. 1
• Although quantum networks include single-sender multiple-receiver broadcast channels, capacity work on entanglement and secret-key distillation with unlimited LOCC had been limited.Ref. [21] established outer bounds for multipartite key agreement and entanglement generation in general single-sender m-receiver channels [21].Source: §I · pp. 1–2
• The paper establishes the unconstrained distillation capacity region for a pure-loss bosonic broadcast channel using state merging for inner bounds and teleportation simulation with relative entropy of entanglement for outer bounds.The bounds match in the infinite-energy limit; determining the finite-energy region remains open.Source: §I · p. 2

II. LOCC-ASSISTED DISTILLATION PROTOCOL AND THE CHANNEL MODEL

• Alice interleaves n channel uses with arbitrary rounds of LOCC to distill entanglement and secret key shared separately with Bob and Charlie.The protocol does not include rates for Bob–Charlie or multipartite entanglement and key.Source: §II · p. 2
• A rate tuple is achievable when protocols exist for every ε in (0, 1) at sufficiently large n, and the capacity region is the closure of all achievable rates.Entanglement rates normalize the logarithm of Schmidt rank by channel uses; secret-key rates normalize the number of key bits.Source: §II · p. 2
• The pure-loss channel divides the input among Bob, Charlie, and the environment with transmittances ηB, ηC, and 1−ηB−ηC, respectively.The transmittances satisfy ηB, ηC ∈ [0, 1] and ηB + ηC ≤ 1; the physical split is modeled with beam splitters and vacuum inputs.Source: §II · p. 2
• The same broadcast channel has multiple beam-splitter implementations, which the paper uses to calculate rate regions.Alternative concatenations first separate different output and environment systems.Source: §II · p. 2

III. UNCONSTRAINED CAPACITY REGION

• Theorem 1 gives the LOCC-assisted, unconstrained capacity region for the pure-loss bosonic broadcast channel.The theorem’s rate expressions are not included in the supplied passage.Source: §III · p. 2
• The appendix generalizes Theorem 1 to the multiple-receiver model from.Source: §III · p. 2

A. Achievability part

• Alice prepares n copies of a two-mode squeezed vacuum and sends A′ through the channel, producing shared copies of φABC.The marginal input mode has average photon number NS.Source: A. Achievability part · p. 3
• Bob and Charlie transfer their systems back to Alice by LOCC using state merging, yielding achievable entanglement rates for Alice with each receiver.The transfer can use successive point-to-point merging or multiparty simultaneous-decoding state merging.Source: A. Achievability part · pp. 2–3
• Because one ebit can generate one private bit, the rate inequalities also apply with EAB and EAC replaced by EAB + KAB and EAC + KAC.Source: A. Achievability part · p. 3
• The right-hand sides of the achievable-rate inequalities are calculated from conditional entropies of the channel output and its purification.The stated entropy difference is g((1−ηC)NS) − g((1−ηB−ηC)NS).Source: A. Achievability part · p. 3
• The achievable rates in (4)–(6) hold in the infinite-NS limit, corresponding to no transmitter energy constraint.Source: A. Achievability part · p. 3

B. Converse part

• Teleportation simulation reduces the adaptive LOCC protocol to distillation of copies of the channel output state, enabling REE bounds.The reduction applies because the pure-loss broadcast channel is covariant under displacement operations.Source: B. Converse part · pp. 3–4
• The converse uses relative entropy of entanglement (REE) to obtain outer bounds on the capacity region.The REE is defined as a minimum over separable states.Source: B. Converse part · pp. 3–4
• −log2(1−η) is the REE of a point-to-point pure-loss channel in the infinite-energy limit.The bound is evaluated using the state obtained by sending an infinite-energy TMSV through the channel.Source: B. Converse part · p. 4
• For the A|BC cut, the REE calculation reduces to a point-to-point pure-loss channel with transmittance ηB + ηC.The second beam splitter is local across the A|BC partition and therefore does not change the REE.Source: B. Converse part · p. 4
• For the C|AB cut, the calculation separates B first, then splits C with transmittance ηC/(1 − ηB), using a Schmidt decomposition and local-unitary invariance of REE.The relevant intermediate thermal state has average photon number (1−ηB)NS.Source: B. Converse part · p. 4
• Figure 2 illustrates the LOCC-assisted capacity region from (4)–(6) for ηB = 0.2 and ηC = 0.3.The passage identifies the plotted region and parameters but does not state a particular boundary comparison.Source: Figure 2 · p. 4, B. Converse part · p. 4

IV. CONCLUSION

• The established region concerns LOCC-assisted bipartite entanglement and secret-key distillation over a pure-loss bosonic broadcast channel.The authors identify it as a potential benchmark for broadcasting these resources.Source: §IV · p. 5
• Extending the scenario to sender–receiver entanglement and key rates, GHZ states, or tripartite common keys remains open, as does constructing protocols that achieve tight inner bounds.An outer bound for the broader scenario was established using squashed entanglement, but the authors note it may be improved.Source: §IV · p. 5
• The finite-energy constrained capacity region remains open; the REE teleportation-reduction bound requires infinite energy for ideal teleportation.The authors point to alternative approaches for tighter outer bounds in some cases.Source: §IV · p. 5
• The REE upper-bound approach could extend to broadcast channels that are teleportation-simulable.The passage frames this as a possible generalization, not an established capacity result for other channels.Source: §IV · p. 5

• The unconstrained LOCC-assisted capacity region is generalized to a pure-loss broadcast channel with arbitrary positive receiver count and transmittances summing to at most one.Source: APPENDIX: THE 1-TO-m BROADCAST CHANNEL · p. 6
• In the infinite-energy limit, the state-merging calculation yields log2([1−ηT]/[1−ηB]); one ebit can generate one private key bit.Source: APPENDIX: THE 1-TO-m BROADCAST CHANNEL · p. 6
• The sequential state-merging protocol is not catalytic: every merging step generates entanglement rather than consuming it.The passage attributes this to the positive value of −H(S1|AS2)φ at every step.Source: APPENDIX: THE 1-TO-m BROADCAST CHANNEL · p. 6
• For the converse, the beam-splitter ordering gives the (t + 1)th splitter transmittance ηT/(1−ηT), which characterizes the REE bound for subset T.Source: APPENDIX: THE 1-TO-m BROADCAST CHANNEL · p. 6, Figure 3 · p. 6