pith. sign in

arxiv: 2604.15733 · v1 · submitted 2026-04-17 · 💻 cs.IT · eess.SP· math.IT

Adaptive Power Allocation and User Scheduling for LEO Satellites using Channel Predictions

Lachlan Drake , Lawrence Ong , Duy T. Ngo This is my paper

Pith reviewed 2026-05-10 08:16 UTC · model grok-4.3

classification 💻 cs.IT eess.SPmath.IT
keywords LEO satellitespower allocationuser schedulingchannel predictionmax-min ratedownlinkuser fairnessoptimization
0
0 comments X

The pith

APASS uses channel predictions to maximize the lowest rate across mobile users in LEO satellite downlinks and keeps fairness near perfect.

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

The paper develops a trajectory-aware channel model that tracks how LEO satellite motion changes link statistics over time for multiple ground users. It then poses a non-convex optimization that maximizes the minimum user rate across a fixed window of time slots by choosing power levels and transmission schedules from predicted future gains. If the approach holds, LEO systems can deliver more balanced service to remote users without needing perfect instantaneous knowledge of every link. Simulations show the resulting scheme stays close to an ideal bound that assumes perfect predictions and raises the worst-case rate nearly threefold over equal-power allocation while holding Jain fairness above 0.99.

Core claim

The authors introduce a channel and transmission model that captures the time-varying statistics caused by satellite motion, formulate the problem of maximizing the minimum rate over multiple slots, and solve it with APASS, which dynamically allocates power and schedules users on the basis of predicted channel gains. Numerical evaluation demonstrates that the scheme remains robust to substantial prediction errors, approaches the performance of an upper bound with perfect foresight, improves the minimum rate by a factor of 2.98 relative to equal-power allocation, and sustains a Jain fairness index above 0.99.

What carries the argument

APASS, the adaptive power allocation and scheduling scheme that solves the max-min rate optimization over a window of time slots by assigning power and transmission slots using predicted future channel gains derived from the trajectory model.

If this is right

  • The minimum rate experienced by the worst-served user rises substantially compared with static equal-power allocation.
  • High user fairness is preserved even when the satellite moves rapidly relative to the ground terminals.
  • The scheme operates close to the performance limit that would be possible with perfect future channel knowledge.
  • Resource allocation remains effective despite sizable errors in the channel predictions supplied to the optimizer.

Where Pith is reading between the lines

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

  • The same prediction-driven max-min formulation could be tested in other fast-moving platforms such as high-altitude platforms or low-orbit constellations with inter-satellite links.
  • Replacing the current predictor with a learned model trained on orbital ephemeris and terrain data might narrow the remaining gap to the perfect-knowledge bound.
  • Hardware-in-the-loop experiments that replay real Doppler and path-loss traces would directly test whether the modeled variability matches observed LEO behavior.

Load-bearing premise

The proposed channel model correctly represents the changes in link statistics caused by the satellite's movement, and sufficiently accurate predictions of future channel gains are available to the optimizer.

What would settle it

Compare the minimum rates and fairness achieved by APASS against measured channel traces collected during an actual LEO satellite pass and check whether the reported factor-of-2.98 gain over equal-power allocation still holds.

Figures

Figures reproduced from arXiv: 2604.15733 by Duy T. Ngo, Lachlan Drake, Lawrence Ong.

Figure 1
Figure 1. Figure 1: Schematic diagram of the APASS successive convex approximation [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (a) The minimum UE rate achieved by APASS as a percentage of the upper bound result and the scheme’s fairness index plotted against the normalized [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
read the original abstract

Low earth orbit (LEO) satellites are a key technology to enable connectivity for rural and remote users. Communication satellites in LEO can provide coverage to much larger areas than terrestrial or aerial systems, while offering improved data rates when compared with geostationary systems. However, a major challenge with LEO satellite communications is the high mobility of the satellite, which results in a rapidly changing communication channel. Due to this, it is challenging to fairly allocate communication resources to multiple users in the system. This work proposes an Adaptive Power Allocation and Scheduling Scheme (APASS) to ensure user fairness in the downlink of a LEO satellite system serving mobile ground users. First, a novel channel and transmission model is introduced to capture the variability in channel statistics due to the satellite's trajectory. Then, a non-convex optimization problem is formulated to maximize the minimum rate across all ground users over a fixed set of time slots. To solve this problem, the proposed APASS dynamically allocates power and schedules transmissions based on predicted future channel gains. Numerical results show that APASS achieves strong performance even with substantial prediction errors, faring close to an upper bound that assumes perfect future channel knowledge. Furthermore, it improves the minimum user rate by a factor of 2.98 compared to equal-power allocation and maintains user fairness with a Jain's fairness index of well above 0.99.

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 / 2 minor

Summary. The manuscript proposes an Adaptive Power Allocation and User Scheduling Scheme (APASS) for the downlink of LEO satellite systems serving mobile ground users. It first introduces a novel channel and transmission model to capture variability in channel statistics induced by the satellite's trajectory. A non-convex optimization problem is then formulated to maximize the minimum rate across users over a fixed set of time slots, solved dynamically by APASS using predicted future channel gains. Numerical simulations are presented showing that APASS achieves a 2.98-fold improvement in minimum user rate over equal-power allocation, maintains a Jain's fairness index well above 0.99, and performs close to an upper bound assuming perfect channel knowledge even under substantial prediction errors.

Significance. If the modeling and simulation assumptions hold, the work provides a practical contribution to fair resource allocation in highly mobile LEO networks, addressing a core challenge for remote connectivity. The explicit formulation of the max-min rate optimization problem and the reported numerical outcomes (including robustness metrics) offer concrete evidence of potential gains over baseline schemes. The approach could support system design if the channel model is shown to generalize.

major comments (2)
  1. [§II] §II (Channel and transmission model): The novel model is load-bearing for all performance claims, yet the manuscript provides no validation against empirical LEO measurements, standard NTN channel models (e.g., 3GPP), or sensitivity analysis to trajectory parameters. Without this, it is unclear whether the reported 2.98× min-rate gain and fairness >0.99 arise from realistic non-stationarity or from an optimistic statistical structure.
  2. [Numerical results section] Numerical results / simulation setup (likely §IV): The Monte-Carlo experiments treat prediction-error variance as a free parameter and claim robustness to 'substantial' errors, but the exact distribution, temporal/spatial correlation structure, and dependence on elevation/user location are not specified. This directly undermines reproducibility and the claim that performance remains close to the perfect-CSI upper bound.
minor comments (2)
  1. [Abstract] The abstract states fairness is 'well above 0.99'; reporting the exact simulated range or minimum value across scenarios would improve precision.
  2. The non-convex optimizer is described as 'dynamically allocates power and schedules transmissions'; a brief outline of the solution method (e.g., successive convex approximation, heuristic, or solver) and its complexity would aid clarity.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive and detailed feedback. We address each major comment point by point below, indicating where revisions will be made to improve clarity, reproducibility, and robustness of the claims.

read point-by-point responses

  1. Referee: [§II] §II (Channel and transmission model): The novel model is load-bearing for all performance claims, yet the manuscript provides no validation against empirical LEO measurements, standard NTN channel models (e.g., 3GPP), or sensitivity analysis to trajectory parameters. Without this, it is unclear whether the reported 2.98× min-rate gain and fairness >0.99 arise from realistic non-stationarity or from an optimistic statistical structure.

    Authors: We agree that the channel model is foundational and that further analysis is warranted. The model is constructed from first principles using the deterministic geometry of LEO satellite motion (time-varying elevation angle, slant range, and resulting path loss) combined with standard Rician fading parameters that vary with elevation. While direct empirical validation against field measurements is outside the scope of this theoretical study, we will add a dedicated paragraph in §II relating the model to existing NTN literature and perform sensitivity analysis by varying key trajectory parameters (orbital altitude 500–1200 km, velocity, and minimum elevation angle). These additions will demonstrate that the reported gains and fairness levels are driven by the adaptive exploitation of non-stationarity rather than any single optimistic assumption. revision: partial

  2. Referee: [Numerical results section] Numerical results / simulation setup (likely §IV): The Monte-Carlo experiments treat prediction-error variance as a free parameter and claim robustness to 'substantial' errors, but the exact distribution, temporal/spatial correlation structure, and dependence on elevation/user location are not specified. This directly undermines reproducibility and the claim that performance remains close to the perfect-CSI upper bound.

    Authors: We acknowledge that the prediction-error model requires explicit specification for reproducibility. In the revised §IV we will state that the error is modeled as additive zero-mean Gaussian noise whose variance is a fixed fraction (10–30 %) of the instantaneous channel gain magnitude, with errors drawn independently across time slots and users. We will also add a brief discussion of possible elevation dependence (larger variance at low elevation) and include additional Monte-Carlo curves that explicitly vary the error variance while keeping all other parameters fixed, thereby confirming that the performance gap to the perfect-CSI bound remains small even under the clarified error model. revision: yes

standing simulated objections not resolved
  • Direct empirical validation of the channel model against real LEO measurement campaigns or full 3GPP NTN channel-model compliance.

Circularity Check

0 steps flagged

No circularity: optimization uses external predictions as inputs

full rationale

The paper defines a novel channel model capturing satellite trajectory effects, then formulates a non-convex optimization problem that takes predicted future channel gains as explicit external inputs and outputs power allocations and schedules to maximize the minimum user rate. Numerical results compare this scheme against equal-power allocation and a perfect-knowledge upper bound via Monte-Carlo simulation; no equation or claim reduces a derived quantity back to a fitted parameter by construction, invokes self-citation for a uniqueness theorem, or renames an input as a prediction. The derivation chain remains self-contained against the stated modeling assumptions and external channel predictions.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on an unstated but load-bearing channel model whose parameters are not enumerated in the abstract, plus the assumption that predictions exist; no new physical entities are introduced.

free parameters (1)
  • prediction error variance
    Numerical robustness claims depend on the magnitude of prediction errors tested in simulation; these levels are chosen but not derived from first principles.
axioms (1)
  • domain assumption Channel statistics vary deterministically with satellite position along its trajectory
    Invoked to justify the 'novel channel and transmission model' that enables the optimization.

pith-pipeline@v0.9.0 · 5548 in / 1446 out tokens · 41986 ms · 2026-05-10T08:16:07.878339+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

23 extracted references · 23 canonical work pages

  1. [1]

    Capacity of the multiple spot beam satellite channel with Rician fading,

    N. Letzepis and A. J. Grant, “Capacity of the multiple spot beam satellite channel with Rician fading,”IEEE Trans. Inf. Theory, vol. 54, Nov. 2008

    work page 2008

  2. [2]

    Massive MIMO transmission for LEO satellite communications,

    L. You, K.-X. Li, J. Wang, X. Gao, X.-G. Xia, and B. Ottersten, “Massive MIMO transmission for LEO satellite communications,”IEEE J. Sel. Areas Commun., vol. 38, pp. 1851–1865, Aug. 2020

    work page 2020

  3. [3]

    Sum rate maximization of massive MIMO NOMA in LEO satellite communication system,

    Z. Gao, A. Liu, C. Han, and X. Liang, “Sum rate maximization of massive MIMO NOMA in LEO satellite communication system,”IEEE Wireless Commun. Lett., vol. 10, no. 8, pp. 1667–1671, 2021

    work page 2021

  4. [4]

    Optimal power control for cognitive LEO constellation with terrestrial networks,

    J. Hu, G. Li, D. Bian, L. Gou, and C. Wang, “Optimal power control for cognitive LEO constellation with terrestrial networks,”IEEE Communications Letters, vol. 24, no. 3, pp. 622–625, 2020

    work page 2020

  5. [5]

    An adaptive power allo- cation for NOMA-based multi-layer satellite system,

    S. Lu, G. Xu, L. Zhu, Z. Song, and W. Zhang, “An adaptive power allo- cation for NOMA-based multi-layer satellite system,”IEEE Transactions on Wireless Communications, vol. 24, no. 6, pp. 5269–5281, 2025

    work page 2025

  6. [6]

    Integrated predictive power control and dynamic channel assignment in mobile radio systems,

    K. Shoarinejad, J. Speyer, and G. Pottie, “Integrated predictive power control and dynamic channel assignment in mobile radio systems,”IEEE Transactions on Wireless Communications, vol. 2, no. 5, pp. 976–988, 2003

    work page 2003

  7. [7]

    3GPP new radio precoding in NGSO satellites: Channel prediction and dynamic resource allocation,

    T. X. Vu, S. Bhandari, M. Minardi, H. Van Nguyen, and S. Chatzinotas, “3GPP new radio precoding in NGSO satellites: Channel prediction and dynamic resource allocation,” in2023 IEEE Statistical Signal Processing Workshop (SSP), pp. 115–119, 2023

    work page 2023

  8. [8]

    Study on new radio (NR) to support non-terrestrial networks,

    “Study on new radio (NR) to support non-terrestrial networks,” TR 38.811, Release 15, 3rd Generation Partnership Project, 2020

    work page 2020

  9. [9]

    Attenuation by atmospheric gases and related effects.,

    “Attenuation by atmospheric gases and related effects.,” ITU-R P.676, International Telecommunication Union, 2022

    work page 2022

  10. [10]

    W. C. Jakes and D. C. Cox,Microwave Mobile Communications. Wiley- IEEE Press, 1994

    work page 1994

  11. [11]

    Two new methods for the generation of multiple uncorrelated Rayleigh fading waveforms,

    M. Patzold and B. Hogstad, “Two new methods for the generation of multiple uncorrelated Rayleigh fading waveforms,” inIEEE VTC, 2006

    work page 2006

  12. [12]

    S. P. Boyd and L. Vandenberghe,Convex optimization. Cambridge University Press, 2004

    work page 2004

  13. [13]

    Power control by geometric programming,

    M. Chianget al., “Power control by geometric programming,”IEEE Trans. Wireless Commun., vol. 6, pp. 2640–2651, July 2007

    work page 2007

  14. [14]

    Deep learning- based channel prediction for LEO satellite massive MIMO communication system,

    Y . Zhang, Y . Wu, A. Liu, X. Xia, T. Pan, and X. Liu, “Deep learning- based channel prediction for LEO satellite massive MIMO communication system,”IEEE Wireless Commun. Lett., vol. 10, Aug. 2021

    work page 2021

  15. [15]

    Channel gain prediction in wireless networks based on spatial-temporal correlation,

    Q. Liao, S. Valentin, and S. Stanczak, “Channel gain prediction in wireless networks based on spatial-temporal correlation,” inIEEE SPA WC, 2015

    work page 2015

  16. [16]

    S. J. W. Jorge Nocedal,Numerical Optimization. Springer New York,

    work page

  17. [17]

    A primal-dual interior-point algorithm for nonsymmetric exponential-cone optimization,

    J. Dahl and E. D. Andersen, “A primal-dual interior-point algorithm for nonsymmetric exponential-cone optimization,”Mathematical Program- ming, 2021

    work page 2021

  18. [18]

    Statistical modeling of the LMS channel,

    F. Fontan, M. Vazquez-Castro, C. Cabado, J. Garcia, and E. Kubista, “Statistical modeling of the LMS channel,”IEEE Transactions on V ehicular Technology, vol. 50, no. 6, pp. 1549–1567, 2001

    work page 2001

  19. [19]

    Deep learning (DL)-based channel prediction and hybrid beamforming for LEO satellite massive MIMO system,

    Y . Zhang, A. Liu, P. Li, and S. Jiang, “Deep learning (DL)-based channel prediction and hybrid beamforming for LEO satellite massive MIMO system,”IEEE Internet of Things Journal, vol. 9, no. 23, pp. 23705– 23715, 2022

    work page 2022

  20. [20]

    Deep learning-based joint channel prediction and multibeam precoding for LEO satellite internet of things,

    M. Ying, X. Chen, Q. Qi, and W. Gerstacker, “Deep learning-based joint channel prediction and multibeam precoding for LEO satellite internet of things,”IEEE Transactions on Wireless Communications, vol. 23, no. 10, pp. 13946–13960, 2024

    work page 2024

  21. [21]

    Parallel forecasting of LEO satellite downlink channels using a CNN-transformer architecture,

    C. Zhao, Y . He, and X. Wang, “Parallel forecasting of LEO satellite downlink channels using a CNN-transformer architecture,” in2025 IEEE 7th International Conference on Communications, Information System and Computer Engineering (CISCE), pp. 97–101, 2025

    work page 2025

  22. [22]

    A quantitative measure of fairness and discrimination for resource allocation in shared computer system,

    R. Jain, D.-M. Chiu, and W. Hawe, “A quantitative measure of fairness and discrimination for resource allocation in shared computer system,” Eastern Research Laboratory, Digital Equipment Corporation, 1984

    work page 1984

  23. [23]

    Proakis and P

    J. Proakis and P. Massoud Salehi,Digital Communications. McGraw-Hill Education, 2007

    work page 2007