Two-Stage Compression of Swarm Position Data for Narrowband Satellite IoT Backhaul

Agent swarms report position continuously. Where no terrestrial network reaches, those reports cross a satellite Internet of Things (IoT) link that carries a few hundred bits per second, far below the rate a swarm produces, so positions arrive late, downsampled, or not at all. We propose a two-stage scheme. The central idea is to divide swarm motion into two levels: common motion shared across all agents and individual perturbations. The separation holds because swarm motion is approximately low-rank under a bounded deformation. Each agent fits K spline coefficients to its own trajectory and sends them, cutting intra-swarm traffic by 92%. At the aggregation node, truncated singular value decomposition (SVD) is applied to the K × M coefficient matrix of the transmitted coefficients. Significant perturbation residuals are then separated and selectively transmitted. At the ground station, the trajectories are rebuilt from two matrix products. On simulated data with 100 agents, a compression ratio of 20:1 holds the worst-vehicle error at 0.65 m. On the NGSIM I-80 dataset, 50 vehicles reach 1.0 m of reconstruction error at 33:1. At 1 kbps, a 128-frame window at a 1 m accuracy uploads in 17.5 s, inside the 25.6 s it lasts.

Authors

Institutions

Publication Details

Journal
Sensors
Published
2026-11-12
DOI
https://doi.org/10.3390/s26227152
Primary Topic
Satellite Communication Systems
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Two-Stage Compression of Swarm Position Data for Narrowband Satellite IoT Backhaul

Qingsong Li, Yiwei Chen, Fengwu Li, Haisu Zhang et al.
Sensors
Satellite Communication Systems
article

Two-Stage Compression of Swarm Position Data for Narrowband Satellite IoT Backhaul

Qingsong Li, Yiwei Chen, Fengwu Li, Haisu Zhang, Liangqin Yu, Xin Zhou
article en

Abstract

Agent swarms report position continuously. Where no terrestrial network reaches, those reports cross a satellite Internet of Things (IoT) link that carries a few hundred bits per second, far below the rate a swarm produces, so positions arrive late, downsampled, or not at all. We propose a two-stage scheme. The central idea is to divide swarm motion into two levels: common motion shared across all agents and individual perturbations. The separation holds because swarm motion is approximately low-rank under a bounded deformation. Each agent fits K spline coefficients to its own trajectory and sends them, cutting intra-swarm traffic by 92%. At the aggregation node, truncated singular value decomposition (SVD) is applied to the K × M coefficient matrix of the transmitted coefficients. Significant perturbation residuals are then separated and selectively transmitted. At the ground station, the trajectories are rebuilt from two matrix products. On simulated data with 100 agents, a compression ratio of 20:1 holds the worst-vehicle error at 0.65 m. On the NGSIM I-80 dataset, 50 vehicles reach 1.0 m of reconstruction error at 33:1. At 1 kbps, a 128-frame window at a 1 m accuracy uploads in 17.5 s, inside the 25.6 s it lasts.

SensorsVol. 26(22)
National University of Defense Technology (CN)
Industry, innovation and infrastructure
Satellite Communication Systems
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon