Emi Digby
About
My research project
Space Compute as a Service: An Adaptive Runtime Platform for Resilient In-Orbit ApplicationsLow Earth Orbit (LEO) satellite constellations are now well established as an enabler of low-latency, high-throughput internet access from anywhere on Earth. With tens of thousands of LEO satellites currently in orbit, interest is growing with both researchers and satellite operators in a potential shift from internet anywhere on earth to compute anywhere on earth, serving compute capability directly from orbit. LEO networks, however, are highly dynamic in nature due to rapid orbital motion, breaking a core assumption made by most software systems: The user could move, but the server most likely will not.
This work aims to explore the runtime platform support required to serve applications and services from LEO networks, arguing that designs for such a platform’s core mechanisms, including orchestration, scheduling and security, should be driven by satellite mobility as a fundamental principle. We initially present TETHER, a prototype runtime employing live state migration and network continuity mechanisms to hide satellite mobility from users. We then discuss proposed plans to extend the TETHER prototype, integrating a mobility-aware control plane, AI-driven scheduling, and attested, identity-preserving live migration. Finally, we aim to explore how the platform could support distributed microservice-based applications with complex inter-service and user-application SLA requirements.
Supervisors
Low Earth Orbit (LEO) satellite constellations are now well established as an enabler of low-latency, high-throughput internet access from anywhere on Earth. With tens of thousands of LEO satellites currently in orbit, interest is growing with both researchers and satellite operators in a potential shift from internet anywhere on earth to compute anywhere on earth, serving compute capability directly from orbit. LEO networks, however, are highly dynamic in nature due to rapid orbital motion, breaking a core assumption made by most software systems: The user could move, but the server most likely will not.
This work aims to explore the runtime platform support required to serve applications and services from LEO networks, arguing that designs for such a platform’s core mechanisms, including orchestration, scheduling and security, should be driven by satellite mobility as a fundamental principle. We initially present TETHER, a prototype runtime employing live state migration and network continuity mechanisms to hide satellite mobility from users. We then discuss proposed plans to extend the TETHER prototype, integrating a mobility-aware control plane, AI-driven scheduling, and attested, identity-preserving live migration. Finally, we aim to explore how the platform could support distributed microservice-based applications with complex inter-service and user-application SLA requirements.
Publications
While compute and application state for networked services are typically hosted on terrestrial servers, a specific subset of workloads such as multi-party interactive applications can benefit from in-orbit compute, particularly in regions with sparse terrestrial network infrastructure. Hosting stateful services on LEO satellites, however, introduces a fundamental challenge: the satellite serving a user changes every few minutes due to orbital motion, threatening to disrupt users' active connection sessions. This paper presents TETHER, a network abstraction that preserves application-visible connectivity despite continual satellite movement. TETHER combines persistent virtual IPs, anticipatory state migration, and transport continuity mechanisms to transparently relocate application state between satellites before performance degradation occurs. The system leverages Linux Checkpoint/Restore In Userspace (CRIU) for process migration and exploits predictable orbital trajectories to proactively schedule migrations. We explore two migration policies with different trade-offs between latency and overhead. directly-above migrates application state to the satellite closest to the user, minimizing latency at the cost of more frequent migrations. In contrast, edge-of-reachability migrates state to a satellite that can stay under an acceptable latency budget the longest, reducing migration frequency while tolerating higher latency. We evaluate both policies using the xeoverse LEO network emulator. Our results show that anticipatory migration preserves long-lived TCP sessions across satellite transitions and enables applications to trade latency tolerance against migration overhead.