Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Decentralized, Decomposition-Based Observation Scheduling for a Large-Scale Satellite Constellation
Authors: Itai Zilberstein, Ananya Rao, Matthew Salis, Steve Chien
JAIR 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The experiments confirm the efficacy of the approach against baseline algorithms, and we discuss the generality of NSS, GND, and properties of COSP to other domains. We present the Neighborhood Stochastic Search (NSS) algorithm, a decentralized algorithm to effectively solve COSP and other large-scale distributed problems, using decomposition. Empirical results demonstrate the efficacy of our approach on small and large problem instances compared to decentralized and centralized baselines. On small problem instances, we show the gap to optimal solutions, while large problem instances enforce the performance at scale, including run-time and message volume results. |
| Researcher Affiliation | Academia | Itai Zilberstein EMAIL Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91011, USA Ananya Rao EMAIL Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91011, USA Carnegie Mellon University, Pittsburgh, PA 15213, USA Matthew Salis EMAIL Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91011, USA Steve Chien EMAIL Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91011, USA |
| Pseudocode | Yes | Algorithm 1 GND Inter-Neighborhood Delegation Algorithm 2 Greedy Scheduler for agent ai Algorithm 3 Random Scheduler for agent ai Algorithm 4 Squeaky Wheel Optimization Algorithm 5 Neighborhood Stochastic Search for agent ai |
| Open Source Code | No | The paper does not provide an explicit statement about the release of source code, nor does it include any links to code repositories. |
| Open Datasets | Yes | We simulate two constellations inspired by operational low Earth orbit constellations (Planet, 2023). The first constellation is modeled on the Dove constellation from Planet, and we refer to it as Planet. ... Planet (2023). Our constellations. https://www.planet.com/our-constellations. Accessed: 2024-08-12. |
| Dataset Splits | No | The paper describes how problem instances are generated for evaluation (e.g., '50 randomly generated small problem instances', '100 randomly generated large problem instances') but does not specify training/test/validation dataset splits in the conventional sense for a fixed dataset. |
| Hardware Specification | No | The paper states 'Simulations are executed in Java.' but does not provide specific details about the hardware (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper states 'Simulations are executed in Java.' but does not provide specific version numbers for Java or any other key software libraries, frameworks, or solvers used in the experiments. |
| Experiment Setup | Yes | For both NSS and SWO, we set max Iterations 20 and determine convergence using a solution difference threshold of 0.001% satisfaction. ... We set Ο such that the largest neighborhood of satellites is at least 10 times smaller than the total constellation size. In the intra-neighborhood delegation heuristic, ΞΆ, we set Zi such that the resultant ordering is based on latitude, then longitude. An example assignment is Z1 2, Z3 1, and Z2 Z4 0. We then evaluate GND(n) with n 1, 2, 4 and 8. ... We use Pu 0.7 as published by the authors of the BD algorithm (Parjan & Chien, 2023), but mention that changing Pu, had minimal effect on the performance of NSS. |