pith. machine review for the scientific record. sign in

arxiv: 2605.10095 · v1 · submitted 2026-05-11 · 💻 cs.NI · cs.SY· eess.SY

Recognition: 2 theorem links

· Lean Theorem

Learning to Compress and Transmit: Adaptive Rate Control for Semantic Communications over LEO Satellite-to-Ground Links

Jiangtao Luo , Yongyi Ran , Guoliang Xu , Jihua Zhou
Authors on Pith no claims yet

Pith reviewed 2026-05-12 03:28 UTC · model grok-4.3

classification 💻 cs.NI cs.SYeess.SY
keywords semantic communicationsLEO satellite linksadaptive rate controljoint source-channel codingreinforcement learningimage transmissionsatellite-to-ground links
0
0 comments X

The pith

RL agent selects JSCC compression levels to hit 95 percent qualified frames in LEO satellite passes.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper builds a reinforcement learning controller that chooses how much to compress each image before transmission over a LEO satellite-to-ground link. The agent uses a short-term prediction of signal strength to set the channel dimension of a SwinJSCC encoder so that as many frames as possible meet both PSNR and MS-SSIM quality targets inside the brief visibility window. An onboard queue model penalizes both overflow and idle time, keeping packet loss at zero. If the policy works as simulated, satellite downlinks of large image collections become far more reliable than with any fixed compression ratio.

Core claim

Under realistic LEO overpass conditions the learned policy transmits nearly 95 percent of frames at acceptable reconstruction quality while incurring zero packet loss, substantially exceeding the performance of fixed-rate baselines.

What carries the argument

The RL agent that maps predicted SNR and current queue state to a channel-dimension choice for the SwinJSCC encoder, trained to maximize qualified frames while respecting buffer limits.

If this is right

  • Satellite operators can downlink larger volumes of imagery within each short visibility window without sacrificing reconstruction quality.
  • Buffer overflow and underutilization are simultaneously avoided, allowing steady use of the limited transmission slot.
  • Proactive rate selection based on predicted SNR outperforms any static compression setting across varying channel conditions.
  • The same framework can be retrained for different image sensors or quality metrics without redesigning the encoder.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could be combined with on-board camera scheduling to decide which scenes to capture and compress first.
  • If the SNR predictor is replaced by real-time channel estimation, the policy might adapt even faster to unexpected fades.
  • Power savings on the satellite become possible by avoiding over-compression on good passes and under-compression on poor ones.
  • The method raises the question of how to keep the RL agent updated when the satellite's orbit or payload characteristics change.

Load-bearing premise

The SNR predictions are accurate enough and the simulation channel model is close enough to real LEO propagation that the learned policy transfers to flight hardware.

What would settle it

Run the trained policy on a real LEO satellite during multiple overpasses and measure the fraction of frames that meet the PSNR and MS-SSIM thresholds; the claim fails if that fraction drops below roughly 80 percent or if packet loss appears.

Figures

Figures reproduced from arXiv: 2605.10095 by Guoliang Xu, Jiangtao Luo, Jihua Zhou, Yongyi Ran.

Figure 1
Figure 1. Figure 1: System architecture of semantic communnications for multispectral [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Transmit queue model. the satellite’s trajectory. This is fundamentally different from a fading channel, where the amplitude scaling is a random, unknown process. The received complex baseband signal is therefore expressed as: y(t) = p Pr(t) · x(t) + n(t), n(t) ∼ CN (0, N0). In engineering practice, the link performance of the satellite￾to-ground feeder link is preferred to be compactly expressed using the… view at source ↗
Figure 3
Figure 3. Figure 3: The overall architecture of GSL-SwinJSCC for image transfer. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Results of model finetuning. 2) DRL Training: During training, the RL agent operates in a simulated environment where the original images are readily available at the gateway. This allows immediate computation of PSNR and MS-SSIM after each transmission, providing instant reward signals for policy optimization. C. Simulation Results 1) SNR sweep and rate adjust: To emulate a full satellite overpass, the el… view at source ↗
Figure 5
Figure 5. Figure 5: Process of SNR sweep and adjust of target rate at [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Distributions of PSNR (a–d) and MS-SSIM (e–h) for reconstructed images. (a)–(d) show PSNR for [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Comparison of reconstruction quality for the lowest-PSNR image per [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Impact of SNR predict on performance of proposed RL policy. [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
read the original abstract

The bottleneck of satellite-to-ground links poses a major challenge for the timely downlink of massive on-board imagery. This paper studies adaptive image transmission over LEO satellite-to-ground links using joint source-channel coding (JSCC). We propose an RL-based framework that dynamically selects the channel dimension (compression ratio) of a SwinJSCC encoder to maximize the number of received satisfying reconstruction-quality constraints (PSNR and MS-SSIM) within a finite visibility window. The agent leverages SNR prediction to perform proactive rate adaptation and incorporates an on-board transmission-queue model that captures bursty encoding while penalizing both buffer overflow and underutilization. Simulations under realistic overpass conditions show that the proposed policy substantially outperforms fixed-rate baselines, achieving nearly 95% qualified frames with zero packet loss.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes an RL-based adaptive rate control framework for joint source-channel coding (JSCC) image transmission over LEO satellite-to-ground links. Using a SwinJSCC encoder, the agent dynamically selects the channel dimension (i.e., compression ratio) by leveraging SNR predictions for proactive adaptation and an on-board transmission queue model that penalizes buffer overflow and underutilization. The objective is to maximize the number of frames satisfying PSNR and MS-SSIM quality thresholds within each finite visibility window. Simulations under realistic overpass conditions report that the learned policy achieves nearly 95% qualified frames with zero packet loss while substantially outperforming fixed-rate baselines.

Significance. If the simulation results prove robust, the work would represent a meaningful advance in applying reinforcement learning to semantic communications for highly dynamic LEO downlinks, where visibility windows are short and channel conditions vary rapidly. The combination of proactive SNR-based adaptation with an explicit queue model addresses practical constraints in on-board imagery transmission. Credit is due for the end-to-end formulation that jointly considers compression, channel use, and buffer dynamics. However, the claimed performance gains rest entirely on simulation fidelity, limiting immediate significance until the modeling assumptions are validated.

major comments (2)
  1. [Simulations (abstract and §5)] The central performance claim (nearly 95% qualified frames, zero packet loss) is obtained from RL policy simulations that rely on an SNR predictor for proactive rate adaptation and a specific transmission-queue model under 'realistic overpass conditions.' No validation of the channel statistics (fading, Doppler, atmospheric effects) against measured LEO traces, nor any sensitivity analysis to SNR prediction error, is described. This is load-bearing because prediction errors directly alter the state transitions observed by the agent and could eliminate the reported advantage over fixed-rate baselines.
  2. [Proposed RL framework (§4)] The reward function and state representation that incorporate the queue penalties for overflow/underutilization and the exact definition of 'qualified frames' (PSNR/MS-SSIM thresholds) are not fully specified. Without these details, it is impossible to assess whether the 95% figure is reproducible or whether the gains are attributable to the RL policy rather than to favorable modeling choices.
minor comments (1)
  1. [Abstract] The abstract states 'substantially outperforms fixed-rate baselines' but does not name the specific baseline rates or compression ratios used; adding this information would improve clarity.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive and detailed feedback. We address each major comment below, clarifying our approach and indicating the revisions planned for the next version of the manuscript.

read point-by-point responses

  1. Referee: [Simulations (abstract and §5)] The central performance claim (nearly 95% qualified frames, zero packet loss) is obtained from RL policy simulations that rely on an SNR predictor for proactive rate adaptation and a specific transmission-queue model under 'realistic overpass conditions.' No validation of the channel statistics (fading, Doppler, atmospheric effects) against measured LEO traces, nor any sensitivity analysis to SNR prediction error, is described. This is load-bearing because prediction errors directly alter the state transitions observed by the agent and could eliminate the reported advantage over fixed-rate baselines.

    Authors: We acknowledge that our evaluation relies on simulation with established LEO channel models (Rician fading, Doppler, and atmospheric effects drawn from prior literature) rather than direct comparison to proprietary measured traces. To directly address the robustness concern, we will add a dedicated sensitivity analysis subsection in §5 that evaluates policy performance under varying levels of SNR prediction error (including zero, moderate, and high error variances). This will quantify how prediction inaccuracies affect the 95% qualified-frame rate and the advantage over fixed-rate baselines. We note that acquiring and integrating specific measured LEO traces would require external datasets beyond our current simulation framework. revision: partial

  2. Referee: [Proposed RL framework (§4)] The reward function and state representation that incorporate the queue penalties for overflow/underutilization and the exact definition of 'qualified frames' (PSNR/MS-SSIM thresholds) are not fully specified. Without these details, it is impossible to assess whether the 95% figure is reproducible or whether the gains are attributable to the RL policy rather than to favorable modeling choices.

    Authors: We apologize for the insufficient detail in the original presentation. The state vector comprises current queue length, predicted SNR over the next visibility slots, remaining time in the overpass window, and recent frame quality feedback. The reward is a linear combination of (i) the count of qualified frames delivered, (ii) a large negative penalty for buffer overflow, and (iii) a smaller penalty for underutilization to promote efficient channel use. Qualified frames are those satisfying both PSNR ≥ 30 dB and MS-SSIM ≥ 0.9. We will expand §4 with the complete mathematical definitions, exact weighting coefficients, and pseudocode for the reward and state transition to ensure full reproducibility. revision: yes

standing simulated objections not resolved
  • Direct validation of the simulated channel statistics against specific measured LEO satellite traces, which would require access to external proprietary or experimental datasets not available within the current study.

Circularity Check

0 steps flagged

No circularity: performance metrics arise from independent simulation of learned RL policy

full rationale

The paper describes an RL agent that selects SwinJSCC channel dimensions to maximize qualified frames within a visibility window, using SNR prediction and a transmission-queue model. The headline result (nearly 95% qualified frames, zero packet loss) is obtained by running the trained policy forward in a simulated environment under stated overpass conditions and comparing against fixed-rate baselines. No equation or claim reduces by construction to a fitted parameter renamed as prediction, no self-citation supplies a load-bearing uniqueness theorem, and the simulation outputs are not algebraically equivalent to the training objective. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no specific free parameters, axioms, or invented entities are detailed. The approach relies on standard RL and JSCC concepts.

pith-pipeline@v0.9.0 · 5445 in / 1267 out tokens · 55689 ms · 2026-05-12T03:28:47.233681+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

99 extracted references · 99 canonical work pages · 4 internal anchors

  1. [1]

    Generative AI-Enabled Semantic Communication: State-of-the-Art, Applications, and the Way Ahead , year=

    Liang, Chengyang and Li, Dong , journal=. Generative AI-Enabled Semantic Communication: State-of-the-Art, Applications, and the Way Ahead , year=

    work page

  2. [2]

    Wireless Image Transmission Using Deep Source Channel Coding With Attention Modules , year=

    Xu, Jialong and Ai, Bo and Chen, Wei and Yang, Ang and Sun, Peng and Rodrigues, Miguel , journal=. Wireless Image Transmission Using Deep Source Channel Coding With Attention Modules , year=

    work page

  3. [3]

    Zhang, Wenyu and Zhang, Haijun and Ma, Hui and Shao, Hua and Wang, Ning and Leung, Victor C. M. , journal=. Predictive and Adaptive Deep Coding for Wireless Image Transmission in Semantic Communication , year=

    work page

  4. [4]

    Channel-Adaptive Semantic Satellite Communication for Remote Sensing Images , year=

    Li, Zhaoyang and Yang, Qianqian , booktitle=. Channel-Adaptive Semantic Satellite Communication for Remote Sensing Images , year=

    work page

  5. [5]

    Deep Joint Source-Channel Coding for Adaptive Image Transmission Over MIMO Channels , year=

    Wu, Haotian and Shao, Yulin and Bian, Chenghong and Mikolajczyk, Krystian and Gündüz, Deniz , journal=. Deep Joint Source-Channel Coding for Adaptive Image Transmission Over MIMO Channels , year=

    work page

  6. [6]

    Semantic-Aware Jointed Coding and Routing Design in Large-Scale Satellite Networks: A Deep Learning Approach , year=

    Gao, Ronghao and Xu, Yunlai and Li, Han and Zhang, Qinyu and Yang, Zhihua , journal=. Semantic-Aware Jointed Coding and Routing Design in Large-Scale Satellite Networks: A Deep Learning Approach , year=

    work page

  7. [7]

    Satellite Transmission with Joint Source-Channel Coding Under Stochastic Processes , year=

    Guo, Yilan and Zhang, Chiya and Gong, Yuanxiang and He, Chunlong , booktitle=. Satellite Transmission with Joint Source-Channel Coding Under Stochastic Processes , year=

    work page

  8. [8]

    Adaptive Semantic Communication System for High-Quality Remote Sensing Image Transmission in Unstable Wireless Environments , year=

    Tan, Zhangyayu and Liu, Caiping and Xie, Kun and Ouyang, Yudian and Wen, Jigang and Zhang, Guangxing and Chen, Dong and Xie, Gaogang and Li, Kenli , journal=. Adaptive Semantic Communication System for High-Quality Remote Sensing Image Transmission in Unstable Wireless Environments , year=

    work page

  9. [9]

    High Quality Compression and Transmission of Remote Sensing Images Based on Semantic Communication , year=

    Jiang, Yan and Xie, Kun and Ouyang, Yudian and Wen, Jigang and Zhang, Guangxing and Liang, Wei and Feng, Quan , journal=. High Quality Compression and Transmission of Remote Sensing Images Based on Semantic Communication , year=

    work page

  10. [10]

    Joint Coding-Modulation for Digital Semantic Communications via Variational Autoencoder , year=

    Bo, Yufei and Duan, Yiheng and Shao, Shuo and Tao, Meixia , journal=. Joint Coding-Modulation for Digital Semantic Communications via Variational Autoencoder , year=

    work page

  11. [11]

    Joint Source and Channel Coding for Multi-Modal Satellite-to-Ground Semantic Communications , year=

    Yin, Yanbo and Liu, Shu and Wen, Dingzhu and Wu, Youlong and Shi, Yuanming , booktitle=. Joint Source and Channel Coding for Multi-Modal Satellite-to-Ground Semantic Communications , year=

    work page

  12. [12]

    Deep Joint Source-Channel Coding for Wireless Image Transmission , year=

    Bourtsoulatze, Eirina and Burth Kurka, David and Gündüz, Deniz , journal=. Deep Joint Source-Channel Coding for Wireless Image Transmission , year=

    work page

  13. [13]

    Toward Wisdom-Evolutionary and Primitive-Concise 6G: A New Paradigm of Semantic Communication Networks , journal =

    Ping Zhang and Wenjun Xu and Hui Gao and Kai Niu and Xiaodong Xu and Xiaoqi Qin and Caixia Yuan and Zhijin Qin and Haitao Zhao and Jibo Wei and Fangwei Zhang , keywords =. Toward Wisdom-Evolutionary and Primitive-Concise 6G: A New Paradigm of Semantic Communication Networks , journal =. 2022 , issn =. doi:https://doi.org/10.1016/j.eng.2021.11.003 , url =

    work page doi:10.1016/j.eng.2021.11.003 2022

  14. [14]

    ComAI: The Convergence of Communication and Artificial Intelligence , year=

    Zhang, Ping and Niu, Kai and Wang, Xiaoyun and Liu, Yiming and Liang, Zijian and Dong, Chen and Dai, Jincheng and Xu, Xiaodong and Xu, Wenjun and Zhang, Zhi and Wang, Guangyu and Li, Yanlu and Wu, Di and Wu, Hequan , journal=. ComAI: The Convergence of Communication and Artificial Intelligence , year=

    work page

  15. [15]

    Proceedings of the ACM SIGCOMM 2025 Conference , pages =

    Sun, Chuanhao and Zhang, Yu and Tao, Bill and Vasisht, Deepak and Marina, Mahesh , title =. Proceedings of the ACM SIGCOMM 2025 Conference , pages =. 2025 , isbn =. doi:10.1145/3718958.3750523 , abstract =

    work page doi:10.1145/3718958.3750523 2025

  16. [16]

    Faster, better, and more accurate mapping of burned areas using Sentinel-2 multispectral images , journal =

    Peng Liu and Yongxue Liu and Xiaoxiao Guo and Yuchen Liu and Wanjing Zhao and Wenxuan Xu , keywords =. Faster, better, and more accurate mapping of burned areas using Sentinel-2 multispectral images , journal =. 2026 , issn =. doi:https://doi.org/10.1016/j.rse.2025.115137 , url =

    work page doi:10.1016/j.rse.2025.115137 2026

  17. [17]

    Sentinel-2 Wildfire Change Detection (S2-WCD) , year =

    Dimitris Valsamis and Alexandros Oikonomidis and Chrysoula Chatzichristaki and Anastasia Moumtzidou and Ilias Gialampoukidis and Stefanos Vrochidis and Ioannis Kompatsiaris , publisher =. Sentinel-2 Wildfire Change Detection (S2-WCD) , year =. doi:10.21227/2t8j-t191 , url =

    work page doi:10.21227/2t8j-t191

  18. [18]

    2025 , eprint=

    Generative Semantic Communication for Joint Image Transmission and Segmentation , author=. 2025 , eprint=

    work page 2025

  19. [19]

    Robust Multimodal Information Bottleneck for Satellite-to-Ground Task-Oriented Communication , year=

    Huang, Jiayi and Wen, Dingzhu and Wu, Youlong and Shi, Yuanming and Wang, Ting , booktitle=. Robust Multimodal Information Bottleneck for Satellite-to-Ground Task-Oriented Communication , year=

    work page

  20. [20]

    Task-Oriented Communication and Optimization Framework for 6G Non-Terrestrial Networks: Challenges and Solutions , year=

    Pei, Jiaming and Dai, Minghui and Al-Dulaimi, Anwer and Al-Rubaye, Saba and Mumtaz, Shahid , journal=. Task-Oriented Communication and Optimization Framework for 6G Non-Terrestrial Networks: Challenges and Solutions , year=

    work page

  21. [21]

    R-ACP: Real-Time Adaptive Collaborative Perception Leveraging Robust Task-Oriented Communications , year=

    Fang, Zhengru and Wang, Jingjing and Ma, Yanan and Tao, Yihang and Deng, Yiqin and Chen, Xianhao and Fang, Yuguang , journal=. R-ACP: Real-Time Adaptive Collaborative Perception Leveraging Robust Task-Oriented Communications , year=

    work page

  22. [22]

    Edge-Cloud Collaborated Object Detection via Bandwidth Adaptive Difficult-Case Discriminator , year=

    Cao, Zhiqiang and Cheng, Yun and Zhou, Zimu and Chen, Yongrui and Hu, Youbing and Lu, Anqi and Liu, Jie and Li, Zhijun , journal=. Edge-Cloud Collaborated Object Detection via Bandwidth Adaptive Difficult-Case Discriminator , year=

    work page

  23. [23]

    Aligning Task- and Reconstruction-Oriented Communications for Edge Intelligence , year=

    Diao, Yufeng and Zhang, Yichi and She, Changyang and Zhao, Philip Guodong and Liying Li, Emma , journal=. Aligning Task- and Reconstruction-Oriented Communications for Edge Intelligence , year=

    work page

  24. [24]

    CV-Cast: Computer Vision–Oriented Linear Coding and Transmission , year=

    Žádník, Jakub and Kieffer, Michel and Trioux, Anthony and Mäkitalo, Markku and Jääskeläinen, Pekka , journal=. CV-Cast: Computer Vision–Oriented Linear Coding and Transmission , year=

    work page

  25. [25]

    SIMAC: A Semantic-Driven Integrated Multimodal Sensing and Communication Framework , year=

    Peng, Yubo and Xiang, Luping and Yang, Kun and Jiang, Feibo and Wang, Kezhi and Oliver Wu, Dapeng , journal=. SIMAC: A Semantic-Driven Integrated Multimodal Sensing and Communication Framework , year=

    work page

  26. [26]

    TOAST: Task-Oriented Adaptive Semantic Transmission Over Dynamic Wireless Environments , year=

    Yun, Sheng and Pei, Jianhua and Wang, Ping , journal=. TOAST: Task-Oriented Adaptive Semantic Transmission Over Dynamic Wireless Environments , year=

    work page

  27. [27]

    Deep Joint Source-Channel Coding for Multi-Task Network , year=

    Wang, Mengyang and Zhang, Zhicong and Li, Jiahui and Ma, Mengyao and Fan, Xiaopeng , journal=. Deep Joint Source-Channel Coding for Multi-Task Network , year=

    work page

  28. [28]

    SNR-EQ-JSCC: Joint Source-Channel Coding With SNR-Based Embedding and Query , year=

    Zhang, Hongwei and Tao, Meixia , journal=. SNR-EQ-JSCC: Joint Source-Channel Coding With SNR-Based Embedding and Query , year=

    work page

  29. [29]

    MambaJSCC: Adaptive Deep Joint Source–Channel Coding With Generalized State Space Model , year=

    Wu, Tong and Chen, Zhiyong and Tao, Meixia and Sun, Yaping and Xu, Xiaodong and Zhang, Wenjun and Zhang, Ping , journal=. MambaJSCC: Adaptive Deep Joint Source–Channel Coding With Generalized State Space Model , year=

    work page

  30. [30]

    SwinJSCC: Taming Swin Transformer for Deep Joint Source-Channel Coding , year=

    Yang, Ke and Wang, Sixian and Dai, Jincheng and Qin, Xiaoqi and Niu, Kai and Zhang, Ping , journal=. SwinJSCC: Taming Swin Transformer for Deep Joint Source-Channel Coding , year=

    work page

  31. [31]

    IEEE Transactions on Mobile Computing , volume=

    A Game-Theoretical Approach for Distributed Computation Offloading in LEO Satellite-Terrestrial Edge Computing Systems , author=. IEEE Transactions on Mobile Computing , volume=

    work page

  32. [32]

    IEEE Internet of Things Journal , volume=

    Graph Neural Networks and Deep Reinforcement Learning-Based Resource Allocation for V2X Communications , author=. IEEE Internet of Things Journal , volume=

    work page

  33. [33]

    IEEE Transactions on Mobile Computing , volume=

    Routing Optimization With Deep Reinforcement Learning in Knowledge Defined Networking , author=. IEEE Transactions on Mobile Computing , volume=

    work page

  34. [34]

    IEEE Internet of Things Journal , volume=

    Distributed Routing and Data Scheduling in IPNs With GNN-Based Multiagent DRL , author=. IEEE Internet of Things Journal , volume=

    work page

  35. [35]

    IEEE Transactions on Intelligent Transportation Systems , volume=

    Spatiotemporal Factorized Graph Neural Networks for Joint Large-Scale Traffic Prediction and Online Pattern Recognition , author=. IEEE Transactions on Intelligent Transportation Systems , volume=

    work page

  36. [36]

    IEEE Transactions on Intelligent Transportation Systems , volume=

    Travel Time and Weather-Aware Traffic Forecasting in a Conformal Graph Neural Network Framework , author=. IEEE Transactions on Intelligent Transportation Systems , volume=

    work page

  37. [37]

    IEEE Transactions on Wireless Communications , volume=

    ENGNN: A General Edge-Update Empowered GNN Architecture for Radio Resource Management in Wireless Networks , author=. IEEE Transactions on Wireless Communications , volume=

    work page

  38. [38]

    Fully-Distributed Dynamic Packet Routing for

    Ran, Yongyi and Ding, Yajie and Chen, Shuangwu and Lei, Jizhao and Luo, Jiangtao , journal=. Fully-Distributed Dynamic Packet Routing for

    work page

  39. [39]

    One Earth , volume=

    Applications in remote sensing to forest ecology and management , author=. One Earth , volume=. 2020 , publisher=

    work page 2020

  40. [40]

    Insects , volume=

    Application of remote sensing data for locust research and management—A review , author=. Insects , volume=. 2021 , publisher=

    work page 2021

  41. [41]

    Remote sensing of environment , volume=

    Remote sensing for agricultural applications: A meta-review , author=. Remote sensing of environment , volume=. 2020 , publisher=

    work page 2020

  42. [42]

    Inside Planet Labs New Satellite Manufacturing Site. 2018

    work page 2018

  43. [43]

    IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , volume=

    Progress and challenges in intelligent remote sensing satellite systems , author=. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , volume=. 2022 , publisher=

    work page 2022

  44. [44]

    Satellite images show utter devastation from wildfires in Maui. 2023

    work page 2023

  45. [45]

    US Space Force satellite data shines light on mystery of Arctic warming

    Samantha Mathewson. US Space Force satellite data shines light on mystery of Arctic warming. 2024

    work page 2024

  46. [46]

    IEEE Geoscience and Remote Sensing Magazine , volume=

    Advances in hyperspectral image and signal processing: A comprehensive overview of the state of the art , author=. IEEE Geoscience and Remote Sensing Magazine , volume=. 2017 , publisher=

    work page 2017

  47. [47]

    Proceedings of the 29th Annual International Conference on Mobile Computing and Networking , pages=

    Transmitting, fast and slow: Scheduling satellite traffic through space and time , author=. Proceedings of the 29th Annual International Conference on Mobile Computing and Networking , pages=

    work page

  48. [48]

    The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences , volume=

    Comparison of open source compression algorithms on VHR remote sensing images for efficient storage hierarchy , author=. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences , volume=. 2016 , publisher=

    work page 2016

  49. [49]

    Satellites in Global Development

    work page

  50. [50]

    2021 IEEE 29th International Conference on Network Protocols (ICNP) , pages=

    Orbitcast: Exploiting mega-constellations for low-latency earth observation , author=. 2021 IEEE 29th International Conference on Network Protocols (ICNP) , pages=. 2021 , organization=

    work page 2021

  51. [51]

    Remote Sensing , volume=

    A review of image super-resolution approaches based on deep learning and applications in remote sensing , author=. Remote Sensing , volume=. 2022 , publisher=

    work page 2022

  52. [52]

    International Journal of Computer Applications , volume=

    Remote sensing satellite image processing techniques for image classification: a comprehensive survey , author=. International Journal of Computer Applications , volume=. 2017 , publisher=

    work page 2017

  53. [53]

    Remote Sensing of Environment , volume=

    Deep learning in environmental remote sensing: Achievements and challenges , author=. Remote Sensing of Environment , volume=. 2020 , publisher=

    work page 2020

  54. [54]

    Computing-Aware Routing for

    Cao, Jiaqi and Zhang, Shengli and Chen, Qingxia and others , journal=. Computing-Aware Routing for. 2023 , publisher=

    work page 2023

  55. [55]

    2022 Asia Communications and Photonics Conference (ACP) , pages=

    High-Dynamic Transmission Modeling for Laser Inter-Satellite Links (LISLs) , author=. 2022 Asia Communications and Photonics Conference (ACP) , pages=. 2022 , organization=

    work page 2022

  56. [56]

    By 2030, a Network of 300 Satellites to Be Operational — Unraveling the Very Low Orbit Integrated Sensing and Communication Satellite Constellation. 2023

    work page 2030

  57. [57]

    IEEE network , volume=

    Satellite mobile edge computing: Improving QoS of high-speed satellite-terrestrial networks using edge computing techniques , author=. IEEE network , volume=. 2019 , publisher=

    work page 2019

  58. [58]

    IEEE Transactions on Wireless Communications , volume=

    Joint optimization of transmission and computation resources for satellite and high altitude platform assisted edge computing , author=. IEEE Transactions on Wireless Communications , volume=. 2021 , publisher=

    work page 2021

  59. [59]

    IEEE Internet of Things Journal , volume=

    Energy-efficient multiaccess edge computing for terrestrial-satellite Internet of Things , author=. IEEE Internet of Things Journal , volume=. 2021 , publisher=

    work page 2021

  60. [60]

    IEEE Transactions on Intelligent Transportation Systems , volume=

    Computation offloading and resource allocation in MEC-enabled integrated aerial-terrestrial vehicular networks: A reinforcement learning approach , author=. IEEE Transactions on Intelligent Transportation Systems , volume=. 2022 , publisher=

    work page 2022

  61. [61]

    Computation offloading in

    Tang, Qingqing and Fei, Zesong and Li, Bin and Han, Zhu , journal=. Computation offloading in. 2021 , publisher=

    work page 2021

  62. [62]

    IEEE Internet of Things Journal , volume=

    Satellite Edge Computing With Collaborative Computation Offloading: An Intelligent Deep Deterministic Policy Gradient Approach , author=. IEEE Internet of Things Journal , volume=. 2023 , publisher=

    work page 2023

  63. [63]

    Deep Reinforcement Learning Based Load Balancing Routing for

    Zuo, Peiliang and Wang, Chen and Wei, Zhanzhen and others , booktitle=. Deep Reinforcement Learning Based Load Balancing Routing for. 2022 , organization=

    work page 2022

  64. [64]

    Towards spatial location aided fully-distributed dynamic routing for

    Xu, Guoliang and Zhao, Yanyun and Ran, Yongyi and others , booktitle=. Towards spatial location aided fully-distributed dynamic routing for. 2022 , organization=

    work page 2022

  65. [65]

    A weighted semi-distributed routing algorithm for

    Guo, Zhe and Yan, Zheng , journal=. A weighted semi-distributed routing algorithm for. 2015 , publisher=

    work page 2015

  66. [66]

    IEEE Access , volume=

    A distributed survivable routing algorithm for mega-constellations with inclined orbits , author=. IEEE Access , volume=. 2020 , publisher=

    work page 2020

  67. [67]

    Computer Communications , volume=

    Graph-based deep learning for communication networks: A survey , author=. Computer Communications , volume=. 2022 , publisher=

    work page 2022

  68. [68]

    International conference on machine learning , pages=

    Discriminative embeddings of latent variable models for structured data , author=. International conference on machine learning , pages=. 2016 , organization=

    work page 2016

  69. [69]

    Advances in neural information processing systems , volume=

    Inductive representation learning on large graphs , author=. Advances in neural information processing systems , volume=

    work page

  70. [70]

    Semi-Supervised Classification with Graph Convolutional Networks

    Semi-supervised classification with graph convolutional networks , author=. arXiv preprint arXiv:1609.02907 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  71. [71]

    International conference on machine learning , pages=

    Neural message passing for quantum chemistry , author=. International conference on machine learning , pages=. 2017 , organization=

    work page 2017

  72. [72]

    Graph Attention Networks

    Graph attention networks , author=. arXiv preprint arXiv:1710.10903 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  73. [73]

    2021 IEEE Wireless Communications and Networking Conference (WCNC) , pages=

    Learning power control for cellular systems with heterogeneous graph neural network , author=. 2021 IEEE Wireless Communications and Networking Conference (WCNC) , pages=. 2021 , organization=

    work page 2021

  74. [74]

    IEEE Transactions on mobile computing , volume=

    Mobile data traffic prediction by exploiting time-evolving user mobility patterns , author=. IEEE Transactions on mobile computing , volume=. 2021 , publisher=

    work page 2021

  75. [75]

    IEEE Transactions on mobile computing , volume=

    Graph attention spatial-temporal network with collaborative global-local learning for citywide mobile traffic prediction , author=. IEEE Transactions on mobile computing , volume=. 2020 , publisher=

    work page 2020

  76. [76]

    Proceedings of the 2018 Workshop on Big Data Analytics and Machine Learning for Data Communication Networks , pages=

    Learning and generating distributed routing protocols using graph-based deep learning , author=. Proceedings of the 2018 Workshop on Big Data Analytics and Machine Learning for Data Communication Networks , pages=

    work page 2018

  77. [77]

    IEEE Transactions on Mobile Computing , year=

    Routing optimization with deep reinforcement learning in knowledge defined networking , author=. IEEE Transactions on Mobile Computing , year=

    work page

  78. [78]

    IEEE Transactions on Industrial Informatics , volume=

    A graph neural network-based digital twin for network slicing management , author=. IEEE Transactions on Industrial Informatics , volume=. 2020 , publisher=

    work page 2020

  79. [79]

    Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , pages=

    Deepwalk: Online learning of social representations , author=. Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , pages=

    work page

  80. [80]

    OPSPF: orbit prediction shortest path first routing for resilient

    Pan, Tian and Huang, Tao and Li, Xingchen and others, Yunjie , booktitle=. OPSPF: orbit prediction shortest path first routing for resilient. 2019 , organization=

    work page 2019

Showing first 80 references.