In emerging scenarios such as the Internet of Things and space-and-ground integrated networks, diverse new applications bring multi-di-mensional quality of service (QoS) requirements to space information networks (SINs). As a mixed problem of routing, traffic control, and resource allocation, QoS provisioning is concentrated on providing customized QoS guarantee to SIN users. On the other hand, motivated by the recent progress in technologies such as software-defined networking and machine learning, many potential solutions for QoS provisioning have been proposed. To utilize the state-of-the-art approaches within QoS provisioning, we provide a forward-looking vision of QoS provisioning that outlines the applications, challenges, architectures, and associated solutions. In this article, we first introduce the applications in three emerging scenarios in terms of traffic characteristics and QoS requirements and identify the accompanying technical challenges for SINs. Then a QoS provisioning architecture with two differentiated transmission modes is proposed. The mirror source mode for bandwidth-tolerant flows improves the network efficiency by centralized traffic scheduling. The preemptive mode for bandwidth-sensitive flows reduces the end-to-end delay by edge node control. Furthermore, the enabling technologies and potential solutions for control logic and forwarding mechanism are analyzed. At last, future research directions are discussed.

中文翻译：

---

空间信息网络中的 QoS 配置：应用、挑战、架构和解决方案


在物联网、天地一体化网络等新兴场景下，多样化的新应用给空间信息网络（SIN）带来了多维度的服务质量（QoS）要求。作为路由、流量控制和资源分配的混合问题，QoS供应集中于为SIN用户提供定制的QoS保证。另一方面，受软件定义网络和机器学习等技术最新进展的推动，人们提出了许多潜在的 QoS 供应解决方案。为了在 QoS 配置中利用最先进的方法，我们提供了 QoS 配置的前瞻性愿景，概述了应用程序、挑战、架构和相关解决方案。在本文中，我们首先从流量特征和 QoS 要求方面介绍了三种新兴场景的应用，并确定了 SIN 所面临的技术挑战。然后提出了一种具有两种不同传输模式的QoS提供架构。带宽容忍流的镜像源模式通过集中流量调度提高网络效率。带宽敏感流的抢占模式通过边缘节点控制减少了端到端延迟。此外，还分析了控制逻辑和转发机制的使能技术和潜在解决方案。最后，讨论了未来的研究方向。