Back to search

Georgia Institute of Technology

Distributed Scheduling for Value and Hardware-Adaptive Satellite Intelligence

Abstract

dc:description.abstract

Earth-observation satellites are gaining popularity as distributed edge computing platforms for latency-sensitive tasks such as disaster response, carrying onboard accelerators that enable orbital inference. However, they face strict energy limits tied to solar cycles. Wasting battery on low-priority tasks can prevent processing highly valuable data during eclipses. Current schedulers rely on rigid priorities or static cutoffs, struggling to adapt to these evolving physical constraints. This thesis introduces Equinox, a decentralized runtime designed for the constraints of the orbital edge. Equinox reduces hardware metrics, including battery charge, thermals, and queue congestion, into a single unified marginal execution cost. Using barrier functions near safety limits, this dynamic cost reflects both immediate and long-term depletion risks. Tasks execute only if their expected scientific value exceeds this cost, enabling natural load shedding. This cost also acts as an implicit coordination mechanism. When local execution becomes too expensive, Equinox delegates tasks to neighboring satellites via inter-satellite links, redistributing workloads efficiently without heavy routing protocols. Evaluated in a 143-satellite simulation, Equinox achieves up to 20% greater scientific return and significantly healthier battery reserves than priority baselines. By strategically shedding low-value work, the system degrades gracefully under contention, proving that hardware-coupled cost signals effectively align distributed scheduling with overarching mission objectives.

Degree

thesis:*
Name thesis:degree_name
Computer Science, MS
Department thesis:degree_department
Computer Science
Grantor
Georgia Institute of Technology
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Erol, Ansel Kaplan
Advisor dc:contributor.advisor
  • Mahajan, Divya
Committee members dc:contributor.committeemember
  • Romero Llamas, Francisco
  • Padmanabha Iyer, Anand
  • Mukhopadhyay, Saibal
ORCID person:identifier.orcid

Rights

Language dc:language.iso
English

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/1853/81643

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Erol, Ansel Kaplan. Distributed Scheduling for Value and Hardware-Adaptive Satellite Intelligence. Georgia Institute of Technology, 2026. https://hdl.handle.net/1853/81643