Optimizing Tree Encodings in Heterogeneous One-way Quantum Repeater Architectures

Large scale quantum communication networks demand flexible and adaptable repeater architectures capable of accommodating physical deployment constraints. We consider an architecture composed of one-way quantum repeaters, where each transmitted logical qubit is encoded into a highly loss-tolerant photonic tree-cluster state. The photonic trees are generated and information is encoded into them using single photon emitters and ancillary qubits coupled together. Given such a network with unevenly spaced repeaters, we tailor the encoding on a per-link basis so as to optimize the tradeoff between the transmission latency and success probability. We observe that global knowledge of the network configuration enables significantly more efficient communication than tree-encoding optimization based only on local link knowledge. We then extend our objective to optimize the rate of transmission. To this end, we design a scheduling strategy that streamlines the transmission of photonic tree-cluster states while efficiently utilizing the limited number of emitters and ancillae at each repeater. This study provides valuable insights into heterogeneous quantum repeater placements, offering a practical pathway toward resource-efficient third generation quantum repeater networks.