LAST PART¶
LIMITATIONS & FUTURE WORK¶
Hypatia is only the first step in building up research infrastructure for a new breed of networks. It has several under-developed pieces, including some where the sparsity of publicly available information was limiting for us.
Hypatia 只是为新型网络构建研究基础设施的第一步。它有几个尚未开发的部分，其中一些部分由于缺乏公开信息而限制了我们的研究。
• The most under-developed aspect is the radio GS-satellite segment design. It would help to frame more realistic models of the interfaces at both satellites and GSes, and for antenna gain and interference.
• The current model of ISLs is also somewhat simplistic, and it would be useful to model the impact of the Doppler effect on the bandwidth and reliability of ISLs.
• Incorporating a weather model would enable work on reliability and rerouting around bad weather.
• Work on multi-path routing and congestion control will also require some modifications to Hypatia.
• GEO-LEO connectivity, albeit not implemented already, should be straightforward to implement if GEO coverage and minimum elevation constraints are known.
• Simulating constellations with heterogeneous satellite and ISL capabilities could be interesting as well — as satellites are gradually deployed, their capabilities may advance over time. Heterogeneity in terms of link capacities is easy to accommodate, but changes to support different numbers of ISLs across satellites will require additional work in defining topologies that appropriately use such heterogeneity.
- 卫星地面站段设计：目前最欠发展的方面是无线电GS-卫星段的设计。构建更现实的卫星和地面站接口模型，以及天线增益和干扰模型将是有帮助的。
- 星间链路（ISL）模型：当前的ISL模型也显得有些简单，建模多普勒效应对ISL带宽和可靠性的影响将是有益的。
- 天气模型：引入天气模型将有助于提高可靠性，并在恶劣天气情况下进行重新路由。
- 多路径路由和拥塞控制：对多路径路由和拥塞控制的研究也需要对Hypatia进行一些修改。
- GEO-LEO连接性：尽管尚未实施，但如果已知GEO覆盖和最低仰角约束，GEO-LEO连接的实现应该是直接的。
- 异构卫星和ISL能力的仿真：仿真具有异构卫星和ISL能力的星座也可能很有趣——随着卫星逐步部署，其能力可能会随着时间而提升。在链路容量方面，异构性容易适应，但支持不同数量的ISL跨卫星的变化将需要额外的工作，以定义适当利用这种异构性的拓扑结构。
More broadly, as is typical for simulation infrastructure, we cannot fully anticipate the needs of novel proposals for LEO networking, and it is likely that many such efforts will require modifications of Hypatia. However, we believe it provides a useful starting point for such work.
由于低地球轨道（LEO）网络的创新提案通常难以完全预见其需求，因此Hypatia可能需要进行多次修改以适应这些新兴需求。然而，我们相信Hypatia为此类工作的开展提供了一个有用的起点。
Importantly, all the takeaways we have highlighted throughout the paper are robust to Hypatia’s current limitations. Regardless of how the missing design details are filled in as more information becomes available, the RTTs are going to vary, congestion control is going to see noisy loss and delay signals, and the shifting paths pose clear challenges for routing and traffic engineering.
重要的是，我们在整篇论文中强调的所有结论都对Hypatia当前的局限性是稳健的 (很难改变这些局限性)。无论随着更多信息的可用，缺失的设计细节如何填补，往返时间（RTT）都会变化，拥塞控制也将面临噪声丢包和延迟信号，而路径的变化则为路由和流量工程带来了明显挑战。
CONCLUSION¶
We present Hypatia, a framework for simulating and visualizing large LEO networks. We demonstrate Hypatia’s utility in understanding the behavior of such networks, especially the temporal variations in the structure of paths and their latencies. We draw out some of the implications of this LEO network dynamism for congestion control, and routing and traffic engineering. Our work not only adds quantitative support to recent position papers drawing out the challenges of LEO networking, but also provides a sorely missing infrastructure for making progress on these challenges.
我们介绍了Hypatia，这是一个用于模拟和可视化大型低地球轨道（LEO）网络的框架。我们展示了Hypatia在理解此类网络行为方面的实用性，特别是路径结构及其延迟的时间变化。我们指出了这种LEO网络动态性对拥塞控制、路由和流量工程的一些影响。我们的工作不仅为最近关于LEO网络挑战的立场文件提供了定量支持，还提供了一个急需的基础设施，以便在这些挑战上取得进展。
ACKNOWLEDGMENTS¶
We are grateful to our shepherd Eric Wustrow and the anonymous reviewers for their helpful feedback.