CONCLUSION

This paper presents STAR FRONT, a pervasive and low-latency framework cooperatively and judiciously exploiting LEO and terrestrial caches for Internet content distribution on a global scale. We implement a testbed to mimic the satellite-cloud integrated architecture and build a STAR FRONT prototype upon it. Trace-driven evaluations show that by cooperatively and properly placing replicas on satellites and clouds, STAR FRONT can satisfy various latency requirements from applications with acceptable cost, as compared to other state-of-the-art cloud-based approaches.

本文介绍了 STAR FRONT，这是一种普及且低延迟的框架，它以协作和明智地利用 LEO 和地面缓存在全球范围内分发互联网内容。我们实施了一个测试平台来模拟卫星云集成架构，并在其上构建了一个 STAR FRONT 原型。跟踪驱动的评估表明，与其他最先进的基于云的方法相比，通过协作和正确地将副本放置在卫星和云上，STAR FRONT 可以以可接受的成本满足应用程序的各种延迟要求。