Practical distributed quantum information processing with LOCCNet

Abstract Distributed quantum information processing is essential for building networks and enabling more extensive computations. In this regime, several spatially separated parties share a multipartite system, the most natural set of operations Local Operations Classical Communication (LOCC). As pivotal part in theory practice, LOCC has led to many vital protocols such as teleportation. However, designing practical challenging due LOCC’s intractable structure limitations by near-term devices. Here we introduce LOCCNet, machine learning framework facilitating protocol design optimization distributed tasks. applications, explore various tasks entanglement distillation, state discrimination, channel simulation. We discover with evident improvements, particular, distillation states interest information. Our approach opens up new opportunities exploring its applications learning, which will potentially sharpen our understanding power LOCC. An implementation LOCCNet available Paddle Quantum, Python package based on PaddlePaddle deep platform.