In this article, a space–air–ground quantum (SPARQ) network is developed as a means for providing a seamless on-demand entanglement distribution. The node mobility in SPARQ poses significant challenges to entanglement routing. Existing quantum routing algorithms focus on stationary ground nodes and utilize link distance as an optimality metric, which is unrealistic for dynamic systems, like SPARQ. Moreover, in contrast to the prior art that assumes homogeneous nodes, SPARQ encompasses heterogeneous nodes with different functionalities further complicates the entanglement distribution. To solve the entanglement routing problem, a deep reinforcement learning (RL) framework is proposed and trained using deep Q-network (DQN) on multiple graphs of SPARQ to account for the network dynamics. Subsequently, an entanglement distribution policy, third-party entanglement distribution (TPED), is proposed to establish entanglement between communication parties. A realistic quantum network simulator is designed for performance evaluation. Simulation results show that the TPED policy improves entanglement fidelity by 3% and reduces memory consumption by 50% compared with benchmark. The results also show that the proposed DQN algorithm improves the number of resolved teleportation requests by 39% compared with shortest path baseline and the entanglement fidelity by 2% compared with an RL algorithm that is based on long short-term memory. It also improved entanglement fidelity by 6% and 9% compared with state-of-the-art benchmarks. Moreover, the entanglement fidelity is improved by 15% compared with DQN trained on a snapshot of SPARQ. Additionally, SPARQ enhances the average entanglement fidelity by 23.5% compared with existing networks spanning only space and ground layers.
In this article, a space–air–ground quantum (SPARQ) network is developed as a means for providing a seamless on-demand entanglement distribution. The node mobility in SPARQ poses significant challenges to entanglement routing. Existing quantum routing algorithms focus on stationary ground nodes and utilize link distance as an optimality metric, which is unrealistic for dynamic systems, like SPARQ. Moreover, in contrast to the prior art that assumes homogeneous nodes, SPARQ encompasses heterogeneous nodes with different functionalities further complicates the entanglement distribution. To solve the entanglement routing problem, a deep reinforcement learning (RL) framework is proposed and trained using deep Q-network (DQN) on multiple graphs of SPARQ to account for the network dynamics. Subsequently, an entanglement distribution policy, third-party entanglement distribution (TPED), is proposed to establish entanglement between communication parties. A realistic quantum network simulator is designed for performance evaluation. Simulation results show that the TPED policy improves entanglement fidelity by 3% and reduces memory consumption by 50% compared with benchmark. The results also show that the proposed DQN algorithm improves the number of resolved teleportation requests by 39% compared with shortest path baseline and the entanglement fidelity by 2% compared with an RL algorithm that is based on long short-term memory. It also improved entanglement fidelity by 6% and 9% compared with state-of-the-art benchmarks. Moreover, the entanglement fidelity is improved by 15% compared with DQN trained on a snapshot of SPARQ. Additionally, SPARQ enhances the average entanglement fidelity by 23.5% compared with existing networks spanning only space and ground layers.
In this article, we present SPARQ: A SPace- Air-gRound Quantum network that supports longrange quantum teleportation and superdense coding in the quantum Internet. SPARQ integrates satellites (space layer), unmanned aerial vehicles (air layer), and remote end users (ground layer) to perform end-to-end entanglement distribution. We first present the architecture of SPARQ, followed by a discussion of its potential applications as well as the challenges facing SPARQ. Next, we investigate one of SPARQ's challenges, the need for efficient routing algorithms. This challenge is particularly significant for SPARQ because of the presence of three layers with diverse distances and capabilities among the nodes, which makes the existing routing algorithms inefficient. Hence, we propose a routing strategy that adopts entanglement fidelity as a routing metric and specifies optimal end-to-end entangled paths while avoiding intensive resource usage. Experimental results demonstrate the superiority of the SPARQ architecture and the proposed routing algorithm. Specifically, SPARQ achieves up to a 5.5% improvement in entanglement fidelity, a 49% increase in entanglement distribution efficiency, and a 10-fold increase in the average number of served requests. The proposed routing algorithm also achieves up to a 7% improvement in entanglement fidelity and a 9-fold increase in the average number of served requests. Finally, we present the conclusion and future directions.
Quantum Internet is essential for unlocking a variety of applications. SPARQ is a space-air-ground quantum network with promising architecture for realizing the quantum Internet. In the early stages of quantum network deployment, limited resources, such as memory units, necessitate efficient management to prevent bottlenecks that can hinder on-demand entanglement distribution. Bottlenecks in quantum networks occur when intermediate repeaters lack the memory needed to accommodate additional entanglement requests. In SPARQ, satellite mobility and fluctuating communication demands lead to continuous changes in bottleneck locations. To address this challenge, in this paper, we propose a proactive entanglement distribution strategy to support on-demand entanglement by predicting and avoiding bottlenecks. Specifically, we propose a graph convolutional neural network (GCNN) framework to predict bottlenecks in SPARQ. The resulting predictions are then integrated into an entanglement distribution policy to proactively avoid bottlenecks, thereby improving the entanglement distribution rate. Simulation results demonstrate the effectiveness of the proposed GCNN model, achieving an accuracy of 97.5% in predicting bottlenecks. The proposed GCNN also improves the accuracy by 6.5% compared to a recurrent neural network (RNN) baseline. This leads to an average improvement of 41.3% in the entanglement rate when avoiding bottlenecks based on GCNN predictions compared with RNN predictions. The results also show that the proposed approach for anticipating bottlenecks leads to a 51.6% improvement in the entanglement distribution rate compared to scenarios where bottlenecks are not addressed.
Efficient routing in entanglement distribution networks is a challenging task that requires the coordinated allocation of quantum memories, repeaters, and quantum operations such as entanglement swapping and distillation across dynamic network conditions. In this paper, we present a Deep Q-Learning (DQN) Two-Stage Routing strategy designed to improve both path and quantum operations selection. Our approach enables the agent to make decisions over a diverse action space, including request scheduling, multipath routing, and the strategic application of quantum operations to maximize entanglement success and resource efficiency. We further extend DQN to integrate Double DQN, Dueling DQN, and Prioritized Experience Replay (PER) architectures. Simulation results indicate that our approach significantly outperforms baseline methods regarding request success rate, especially in large-scale networks under strict fidelity requirements.
Efficient entanglement routing in quantum satellite networks directly affects quantum communication reliability, which is reflected by entanglement fidelity. However, due to dynamic topologies of the satellites and limited quantum resources, the entanglement routing optimization problem remains an open challenge. Classical routing methods are inefficient to solve the problem due to their limited serial processing capability in handling such high computational complexity. To address this problem, this paper proposes a multi-modal quantum deep reinforcement learning (QDRL) model, capable of processing time-varying and complex features. Here, the multi-modal refers to leveraging heterogeneous data from multiple environmental features, rather than relying on homogeneous data, to improve performance. QDRL is then used to optimize the entanglement routing to achieve maximum entanglement fidelity. To evaluate performance, extensive simulations are conducted to compare the proposed multi-modal QDRL framework with a multi-modal classical DRL (CDRL) approach and a homogeneous QDRL scheme that relies on homogeneous data. Simulation results demonstrate that the proposed multi-modal QDRL framework achieves significantly faster convergence, outperforming the CDRL scheme by 99.71% and the homogeneous QDRL scheme by 5.06%. Furthermore, the proposed multi-modal QDRL framework achieved higher entanglement fidelity as compared to the CDRL and homogeneous QDRL schemes by 6.73% and 2.40%, respectively.
Quantum network can provide end-to-end entanglements through quantum operations, supporting various quantum applications such as quantum key distribution and distributed quantum computing. However, many previous proposals for quantum networks assumed the division of quantum network into time slots and batch processing of network communication tasks, performing poorly in dynamic network scenarios and not fully utilizing network resources. This work focuses on the routing problem in entanglement quantum networks, aiming to provide more efficient end-to-end entanglement establishment for upper-layer applications. We propose an on-demand routing algorithm that dynamically adjusts the optimal path based on the current network conditions. Specifically, we consider the benefits of path width for quantum networks and propose an algorithm for finding the shortest wide path. Using this algorithm, each node establishes its own routing table. Additionally, to dynamically avoid network hotspots, we introduce the concept of traffic to reflect the busyness of nodes. Finally, using a dynamic cost calculation method as the routing metric, we select the optimal path for the current request. To evaluate the performance of the proposed quantum routing algorithm, we conducted simulation experiments on SimQN quantum simulation platform. The results show that our approach improves network throughput by 14 percent compared to existing solutions.
Entanglement generation in long-distance quantum networks is challenging because resources are limited and entanglement swapping is probabilistic. To maximize the rate of successful requests, existing quantum routing algorithms often rely on computationally expensive methods such as Integer Linear Programming (ILP) to determine which links to entangle and use for end-to-end entanglement generation. However, these approaches fail to meet the latency requirements of real-world quantum networks. In this study, we propose a Reinforcement Learning (RL)-based model that determines which links to entangle in each time slot, replacing the slow ILP-based link-selection phase used in prior algorithms. The proposed Deep Q-learning model is up to 19.2× faster than linear programming in link-selection while maintaining comparable routing performance. The RL link-selection model alone matches ILP in request success rate; combining RL link selection with entanglement caching and proactive swapping raises throughput by up to 52.55% over ILP. Overall, our approach achieves success rates exceeding state-of-the-art solutions while reducing execution time by more than an order of magnitude. Experiments on a synthetic 50-node Waxman topology and the real-world 54-node SURFnet core topology confirm that these gains generalize across network structures, providing a practical path toward scalable quantum routing.