# LOCCNet: a machine learning framework for distributed quantum information processing

**Authors:** Xuanqiang Zhao, Benchi Zhao, Zihe Wang, Zhixin Song, Xin Wang  
**arXiv:** [2101.12190](https://arxiv.org/abs/2101.12190)  
**Date:** 2021-01-28  
**Summary page:** https://www.summarizepaper.com/en/paper/2101.12190v1/

## Key points

- Distributed quantum information processing is crucial for quantum networks and complex quantum computations
- Local Operations and Classical Communication (LOCC) are natural operations in this framework
- LOCC has led to important protocols like quantum teleportation
- LOCCNet is a machine learning framework for designing and optimizing LOCC protocols
- LOCCNet has applications in entanglement distillation, quantum state discrimination, and quantum channel simulation
- Novel protocols discovered through LOCCNet show significant improvements in entanglement distillation
- LOCCNet enhances our understanding of the power and limitations of LOCC
- Quantum channels are important for characterizing physical implementations of quantum states
- LOCCNet outperforms traditional methods in distinguishing between Bell states and noisy Bell states
- This research advances distributed quantum information processing using machine learning techniques
- It improves our understanding of entanglement and its applications within current technology constraints
- This work paves the way for future advancements in practical quantum networks and extensive quantum computations.

## Summary

Distributed quantum information processing is crucial for the development of quantum networks and enabling more complex quantum computations. In this framework, multiple parties who are physically separated share a multipartite quantum system, and the most natural operations that can be performed are Local Operations and Classical Communication (LOCC). LOCC has played a pivotal role in quantum information theory and practice, leading to important protocols such as quantum teleportation. To address the challenges posed by near-term quantum devices when designing practical LOCC protocols, the authors introduce LOCCNet - a machine learning framework that facilitates the design and optimization of protocols for distributed quantum information processing tasks. They explore various applications of LOCCNet in tasks such as entanglement distillation, quantum state discrimination, and quantum channel simulation. Through their approach they discover novel protocols that demonstrate significant improvements in entanglement distillation with specific types of relevant to quantum information states. The introduction of LOCCNet opens up new opportunities for exploring entanglement and its applications using machine learning techniques. This advancement has the potential to enhance our understanding of the power and limitations of LOCC. Additionally, the authors highlight the importance of studying quantum channels as they characterize what can be achieved with physical implementations of quantum states. In their experiments they evaluate the success probability of distinguishing between a Bell state and a noisy Bell state using both traditional protocols and those designed through LOCCNet. The results show that the protocol implemented via LOCCNet outperforms traditional methods. Overall, this research contributes to advancing distributed quantum information processing by leveraging machine learning techniques to design more efficient protocols. By improving our understanding of entanglement and its applications within the constraints imposed by current technology, this work paves the way for future advancements in building practical quantum networks and performing extensive quantum computations.