routing

This paper addresses the performance trade-offs of routing strategies in large-scale, highly dynamic satellite networks. It systematically compares centralized versus distributed routing with respect to throughput and end-to-end latency, focusing on scenarios where network state change rates approach signal propagation/transmission delays—and under both onboard buffer availability and absence. Leveraging a shortest-path–based routing model, the study integrates link-state freshness analysis with theoretical performance derivation. The work provides the first quantitative proof that, in buffer-less scenarios, distributed routing significantly improves throughput; whereas, when onboard buffers are present, it reduces end-to-end latency while maintaining equivalent throughput. These findings offer rigorous theoretical foundations and design guidelines for latency-critical low-Earth-orbit (LEO) satellite communications.

To address the path optimization challenge for dynamic traffic engineering in software-defined networking (SDN), this paper proposes a real-time closed-loop control framework integrating deep reinforcement learning (DRL) with source routing. Methodologically, we design PolKA—a lightweight, P4-programmable source routing mechanism—and Hecate—a DRL-based system for real-time traffic analytics and path decision-making—achieving, for the first time, their coordinated closed-loop scheduling on a physical P4 testbed. Our key contribution is a data-plane-aware source routing integration paradigm that tightly couples path intelligence with programmable forwarding. Experimental results demonstrate a 37% reduction in end-to-end scheduling latency and a 52% decrease in link utilization variance, significantly enhancing network adaptability and operational controllability.

Low Earth Orbit (LEO) satellite networks face severe resilience challenges due to frequent, unpredictable link and node failures—including those induced by cyberattacks—compromising routing reliability. Method: This paper proposes and systematically compares three fault-aware rerouting paradigms: local-neighbor-based, segment-based, and global-knowledge-based approaches. Using an extended Deep Space Network Simulator (DSNS), we quantitatively evaluate delivery ratio, end-to-end latency, rerouting overhead, and loop occurrence under both random and targeted failure scenarios. Results: Segment-based rerouting achieves the optimal trade-off between responsiveness and coordination cost: it significantly improves fault tolerance over local strategies while reducing communication and computational overhead by an order of magnitude compared to global strategies—all while maintaining high delivery ratio and low latency. This work is the first to establish segment-awareness as a fundamental design principle balancing scalability and survivability, thereby providing both theoretical foundations and practical routing paradigms for resilient large-scale LEO networks.

With the rapid growth in data demand and stringent latency requirements of modern applications has driven significant interest in Low Earth Orbit (LEO) satellite constellations as an emerging solution for global Internet coverage. However, routing in LEO networks remains a fundamental challenge due to highly dynamic topologies, time-varying traffic conditions, and its susceptibility to link failures. Conventional routing algorithms typically assume static link metrics and fail to account for queue backlogs or real-time system variations, making them less effective in such environments. We propose a queue-aware multi-agent deep reinforcement learning (MA-DRL) framework for routing in LEO satellite networks. Each satellite is modeled as an independent agent responsible for making local routing decisions, enabling a distributed and scalable solution. The proposed framework formulates a latency-aware optimization problem that incorporates background traffic, queue dynamics at each satellite, and a resilience score to improve robustness. We evaluate the proposed approach against the state-action-reward-state-action (SARSA) and Dijkstra algorithms. While Dijkstra achieves the lowest end-to-end latency under ideal conditions, its computational and signaling overhead becomes a significant bottleneck as the network scales. In contrast, our proposed approach incurs significantly lower overhead (approximately 50% of Dijkstra at a 5 s recalculation interval), scales efficiently with network size, and effectively manages queue backlogs and resilience under increasing traffic load, demonstrating enhanced robustness and scalability in LEO satellite networks while maintaining competitive latency and resilience scores.

This study addresses the scalability bottleneck in unicast and multicast routing caused by the dual role of IP addresses as both identifiers and locators. It systematically traces the evolution of Internet routing scalability solutions, first articulating the map-and-encap architecture as a unifying paradigm and identifying the essential conditions for its successful deployment. Through historical protocol analysis, architectural comparisons, and conceptual abstraction—encompassing approaches such as BIER and tunnel encapsulation—the work reveals that BGP’s lack of intra-domain egress router topology abstraction is a fundamental limitation. The paper proposes core principles to guide future scalable routing designs, emphasizing the critical roles of locally driven incentives and effective topology abstraction in protocol evolution.

Existing load-balancing mechanisms struggle to effectively exploit path diversity in low-diameter topologies such as Dragonfly and Slim Fly, often relying on proprietary hardware or lacking adaptivity. This work proposes Spritz, a sender-based, general-purpose Ethernet load-balancing framework that, for the first time, enables topology-aware adaptive routing without requiring additional hardware support. Spritz integrates two complementary algorithms—Spritz-Scout and Spritz-Spray—that leverage ECN feedback, packet truncation, and timeout signals for efficient path probing and selection, augmented by a caching mechanism to enhance performance. Large-scale simulations at the thousand-node level demonstrate that Spritz reduces flow completion times by up to 1.8× compared to ECMP and UGAL-L under normal conditions, and achieves up to a 25.4× improvement in the presence of link failures.

This study addresses energy efficiency optimization in communication networks during low-traffic periods by jointly optimizing network topology design and shortest-path routing. The approach ensures that all traffic demands can be satisfied within the activated subnetwork through dynamically adapted shortest paths. The authors propose, for the first time, a capacitated integer linear programming model that precisely captures dynamic shortest-path routing, complemented by provably effective strengthening constraints to accelerate solution convergence. A tailored column generation algorithm is developed to efficiently handle large-scale instances. Experimental results demonstrate that a simplified strategy—fixing routes and deactivating redundant links—achieves near-optimal performance, while the traffic-oblivious method TOCA exhibits superior efficacy in multi-demand scenarios.

This work proposes a predictive traffic scheduling framework based on Network Digital Twin (NDT) to overcome the limitations of traditional routing protocols, which react passively only after performance degradation and struggle to handle congestion caused by dynamic traffic and topological changes. The framework uniquely integrates graph generative models—Erdős–Rényi, Barabási–Albert, and Watts–Strogatz—with Message Passing Neural Networks (MPNNs) to continuously mirror and forecast the physical network state in real time. Leveraging a Policy-Based Routing (PBR) feedback mechanism, the approach enables non-intrusive, globally optimized congestion-aware scheduling. Evaluated under synthetic traffic loads, the method accurately classifies link congestion states, significantly improving network throughput and reducing end-to-end latency.

This study addresses the problem of achieving global reachability via greedy routing in networks of autonomous agents equipped only with local views, while minimizing the number of connections per agent. By formulating a strategic link-formation game in a metric space, agents simultaneously minimize their own degrees and ensure network-wide reachability. The work reveals, for the first time, the fundamental mechanisms underlying greedy-routing-compatible network formation without relying on distributional assumptions or predefined protocols, and provides a systematic analysis distinguishing directed and undirected settings. The main contributions include proving that in the directed case, an optimal equilibrium exists and is efficiently computable; in the undirected two-dimensional setting, the proposed approximate equilibrium outperforms Delaunay triangulation, achieves a price of anarchy between 1.75 and 1.8, and is computable in polynomial time.