January 2025 – Sanidhay Bhambay, Thirupathaiah Vasantam, Neil Walton
Contact for further information: Sanidhay Bhambay
This work has been accepted at ACM SIGMETRICS June 2025: Full article
Distributed quantum communication is a fast-growing field with immense importance. It addresses the increasing demand for quantum-secure networks, enabling communication. The potential of distributed quantum computation further drives this progress as researchers and industries work toward realizing the benefits of quantum computation. The Quantum Internet is no longer just a concept; it is becoming a reality, evidenced by large-scale quantum communication networks deployed worldwide. For example, China’s nationwide quantum key distribution (QKD) network spans over 1,120 km, while the UK, Europe, and the US are also advancing their quantum communication infrastructures.
Components such as quantum repeaters and switches are necessary to utilize the potential of distributed quantum communication. These technologies enable entanglement swapping and routing, analogous to the routers in today’s Internet. However, despite the rapid pace of experimental and commercial advancements, no one knows how these distributed systems should work optimally. Fundamental questions remain about their design and what constitutes their optimal performance. We focus on addressing these questions. Our research focuses on the design of optimal distributed quantum switches, laying the groundwork for efficient, programmable, and scalable quantum network infrastructure.
Specifically, the Quantum Internet relies on two fundamental components: quantum repeaters and quantum switches, as shown in the figure above. Repeaters extend the range of quantum communication by distributing entangled particles over long distances, while switches act as central nodes, managing and routing these entangled states among users. The figure illustrates this process, showing users connected to switches or repeaters via entangled links (red lines). Quantum switches manage these entanglements, performing measurement operations to meet user requests.
Our research focuses on designing an optimal quantum switch that efficiently manages user requests and maximizes network resource utilization. By analyzing entanglement distribution and storage dynamics, we characterize what an optimal switch should look like and how it can allocate resources dynamically to handle competing user requests. Here, AI plays an important role. We find that optimal scheduling design requires the solution of an optimization called an average reward Markov Decision Process. These systems are typically solved with AI algorithms. We now begin to analyze these systems and AI algorithms.
One can envisage a future where quantum internet routers implement distributed AI algorithms. These AI-enabled switches might contain an AI chip capable of solving reinforcement learning problems that optimally allocate capacity and distribute entanglement across a quantum network.