pith. sign in

arxiv: 2602.23788 · v2 · pith:SDCH356Inew · submitted 2026-02-27 · 💻 cs.NI

Deep Sleep Scheduling for Satellite IoT via Simulation Based Optimization

Wanja de Sombre , Monika Tomov\'a , Marek Galinski , Anja Klein , Andrea Ortiz This is my paper

Pith reviewed 2026-05-21 11:44 UTC · model grok-4.3

classification 💻 cs.NI
keywords satellite IoTdeep sleepsimulation based optimizationMarkov decision processenergy efficiencydata qualityGoal-Oriented Tensor metric
0
0 comments X

The pith

Probabilistic forecasts let satellite IoT sensors optimize deep-sleep durations for better energy and data trade-offs

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

The paper models satellite IoT connectivity using a Markov decision process to decide when sensors should enter deep sleep. Data quality is assessed with a Goal-Oriented Tensor that considers both the age and the content of received information. The proposed probabilistic simulation-based optimization uses estimates of state transition probabilities to simulate possible futures and pick the sleep length that best balances low energy use with minimal quality loss. This is important for devices in remote locations that rely on satellite links and limited power supplies. Simulations and hardware tests indicate the method works well under changing conditions.

Core claim

The central claim is that probabilistic simulation-based optimization allows a satellite IoT sender to estimate transition probabilities of the observed process and Markov channels online, simulate future states, and select deep-sleep durations that minimize a weighted sum of energy consumption and data quality degradation as measured by the Goal-Oriented Tensor.

What carries the argument

Probabilistic simulation-based optimization (PSBO) that forecasts future states from estimated transition probabilities to determine optimal deep-sleep durations.

If this is right

  • Longer sleep intervals can be used safely when forecasts indicate that data transmission is not urgent.
  • Transmissions can be timed to avoid periods of high delay or erasure predicted by the simulations.
  • The online learning of probabilities enables adaptation to unknown and changing environments.
  • Overall battery life extends while maintaining acceptable data quality at the receiver.

Where Pith is reading between the lines

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

  • The technique may apply to other wireless IoT setups with intermittent links and energy constraints.
  • Future work could combine this with machine learning for even better probability estimates.
  • Field tests in actual satellite environments would confirm if the Markov assumption holds over long periods.

Load-bearing premise

The processes governing data generation and channel behavior are Markovian, permitting reliable online estimation of transition probabilities for state forecasts.

What would settle it

If experiments show that PSBO yields higher total costs than baseline sleep policies when the data process has memory beyond the current state or channels exhibit non-Markovian patterns.

Figures

Figures reproduced from arXiv: 2602.23788 by Andrea Ortiz, Anja Klein, Marek Galinski, Monika Tomov\'a, Wanja de Sombre.

Figure 1
Figure 1. Figure 1: System Model decrease, we use a cap M ∈ N for age-based metrics, meaning that instead of increasing indefinitely, these metrics grow until they reach the value M and stay constant at M afterwards. The GoT metric is defined as a mapping GoT : X × X × {0, . . . , M} → R, (5) where X denotes the space of possible process states, and M is the maximum value of the considered age-based metric F. Notably, widely-… view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the device’s operation schedule. Top: Al [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overview of the interaction between Alg. 1 to 4. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Physical layout of the experimental setup consisting of [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Average cost of PSBO and baselines for varying parameters (GoT [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Here, we assume that the device measures temperatures. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average cost of PSBO and baselines for a varying parameters (GoT [PITH_FULL_IMAGE:figures/full_fig_p012_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Experimental Results for our PSBO approach and baseline strategies [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
read the original abstract

The Satellite Internet of Things (S-IoT) enables global connectivity for remote sensing devices that must operate energy-efficiently over long time spans. We consider an S-IoT system consisting of a sender-receiver pair connected by a data channel and a feedback channel and capture its dynamics using a Markov Decision Process (MDP). To extend battery life, the sender has to decide on deep-sleep durations. Deep-sleep scheduling is the primary lever to reduce energy consumption, since sleeping devices consume only a fraction of their idle power. By choosing its deep-sleep duration online, the sender has to find a trade-off between energy consumption and data quality degradation at the receiver, captured by a weighted sum of costs. We quantify data quality degradation via the recently introduced Goal-Oriented Tensor (GoT) metric, which can take both age and content of delivered data into account. We assume a Markovian observed process and Markov channels with time-varying delay and erasure rates. The challenge is that content awareness of the GoT metric makes periodic transmissions inherently inefficient. Additionally, optimal sleep durations depends on the (unknown) future states of the observed process and the channels, both of which must be inferred online. We propose a novel algorithm using probabilistic simulation-based optimization (PSBO). With PSBO, the sensor forecasts future states based on estimated transition probabilities, and uses these forecasts to select the optimal deep-sleep duration. Extensive simulations and experiments with S-IoT hardware demonstrate superior performance of PSBO under diverse conditions.

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

1 major / 1 minor

Summary. The manuscript presents a deep-sleep scheduling algorithm for Satellite IoT (S-IoT) systems using probabilistic simulation-based optimization (PSBO). The system is modeled as a Markov Decision Process (MDP) with a sender deciding deep-sleep durations to balance energy consumption against data quality degradation, quantified using the Goal-Oriented Tensor (GoT) metric. The approach assumes Markovian observed processes and Markov channels with time-varying delay and erasure rates, estimating transition probabilities online to forecast future states and select optimal sleep durations. Extensive simulations and hardware experiments are claimed to show superior performance under diverse conditions.

Significance. If the PSBO forecasts remain accurate despite time variation and the GoT-based trade-off is shown to be superior in reproducible experiments, the work could advance energy-efficient remote sensing in global S-IoT by providing an online, simulation-driven alternative to static or periodic scheduling. The explicit use of content-aware quality metrics and probabilistic forecasting constitutes a concrete technical contribution.

major comments (1)
  1. Abstract (modeling assumptions paragraph): the claim that PSBO selects optimal deep-sleep durations rests on the ability to generate accurate multi-step forecasts from estimated transition probabilities. However, the same paragraph states that the channels have time-varying delay and erasure rates. If these rates render the underlying chain non-stationary, fixed or slowly updated transition probabilities cannot be expected to produce reliable forecasts; this directly undermines the central optimality and performance claims. A concrete test (e.g., forecast error under controlled non-stationarity) is needed.
minor comments (1)
  1. Abstract: no numerical performance deltas, baseline comparisons, or error bars are supplied despite the assertion of 'superior performance'; adding one or two key quantitative results would make the summary self-contained.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comment on the modeling assumptions. We address the concern regarding non-stationarity and online estimation below, and we will revise the manuscript to strengthen the presentation of forecast reliability.

read point-by-point responses

  1. Referee: Abstract (modeling assumptions paragraph): the claim that PSBO selects optimal deep-sleep durations rests on the ability to generate accurate multi-step forecasts from estimated transition probabilities. However, the same paragraph states that the channels have time-varying delay and erasure rates. If these rates render the underlying chain non-stationary, fixed or slowly updated transition probabilities cannot be expected to produce reliable forecasts; this directly undermines the central optimality and performance claims. A concrete test (e.g., forecast error under controlled non-stationarity) is needed.

    Authors: We appreciate this observation. The manuscript models the channels as Markovian with time-varying parameters, but explicitly states that transition probabilities are estimated online from recent observations to enable adaptive multi-step forecasts. This online procedure updates the estimated model using a sliding window of recent state transitions, allowing it to track changes in delay and erasure rates without assuming stationarity. The optimality claim for PSBO therefore rests on this adaptive forecasting rather than fixed probabilities. To directly address the request for a concrete test, we will add a new simulation experiment in the revised manuscript that measures multi-step forecast error (e.g., mean absolute error on predicted states) under controlled non-stationary channel conditions with varying rates, and we will update the abstract to clarify the online adaptation mechanism. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the PSBO derivation

full rationale

The paper models the S-IoT system as an MDP under Markovian assumptions for the observed process and channels (with time-varying rates noted but not altering the estimation step). PSBO forecasts future states from online-estimated transition probabilities and selects sleep durations via simulation-based optimization. This is an independent algorithmic procedure, not a self-definitional loop or fitted parameter renamed as prediction. Validation comes from separate simulations and hardware experiments, which do not reduce to the modeling inputs by construction. No self-citations, uniqueness theorems, or ansatzes are invoked in a load-bearing way in the provided text. The derivation chain remains self-contained.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the ability to estimate transition probabilities accurately enough for useful forecasts and on the Markov property allowing the MDP formulation.

free parameters (1)
  • weights in the weighted sum of costs
    The trade-off between energy consumption and data quality degradation is expressed as a weighted sum whose specific weights must be chosen or tuned.
axioms (1)
  • domain assumption The observed process and the channels are Markovian with time-varying delay and erasure rates.
    This assumption enables the MDP model and the online estimation of transition probabilities used for forecasting.

pith-pipeline@v0.9.0 · 5813 in / 1325 out tokens · 49388 ms · 2026-05-21T11:44:13.820863+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

28 extracted references · 28 canonical work pages

  1. [1]

    On pursuit of sleep-scheduling for sensor network in IIoT: A non-linear AoI optimization perspective,

    X. Wang, X. Li, M. Zhu, Q. Gao, Y . Huo, and T. Jing, “On pursuit of sleep-scheduling for sensor network in IIoT: A non-linear AoI optimization perspective,” inProc. of 16th International Conference on Communication Software and Networks, 2024

    work page 2024

  2. [2]

    Optimal sleep scheduling for energy-efficient aoi optimization in industrial internet of things,

    X. Cao, J. Wang, Y . Cheng, and J. Jin, “Optimal sleep scheduling for energy-efficient aoi optimization in industrial internet of things,”IEEE Internet of Things Journal, 2023

    work page 2023

  3. [3]

    Sleep, sense or transmit: Energy-age tradeoff for status update with two-threshold optimal policy,

    J. Gong, J. Zhu, X. Chen, X. Ma, and C. M. C., “Sleep, sense or transmit: Energy-age tradeoff for status update with two-threshold optimal policy,” IEEE Transactions on Wireless Communications, 2022

    work page 2022

  4. [4]

    Satellite intermittent connectivity and its impact on age of information for finite horizon scheduling,

    L. Badia and A. Munari, “Satellite intermittent connectivity and its impact on age of information for finite horizon scheduling,” inProc. of 12th Advanced Satellite Multimedia Systems Conference and the 18th Signal Processing for Space Communications Workshop, 2025

    work page 2025

  5. [5]

    Minimizing the age of incorrect information for status update systems with energy harvesting,

    S. Dongare, A. Jovovic, W. de Sombre, A. Ortiz, and A. Klein, “Minimizing the age of incorrect information for status update systems with energy harvesting,” inProc. of IEEE International Conference on Communications, 2024

    work page 2024

  6. [6]

    Minimizing age of incorrect information for unreliable channel with power constraint,

    Y . Chen and A. Ephremides, “Minimizing age of incorrect information for unreliable channel with power constraint,” inProc. of IEEE Global Communications Conference, 2021

    work page 2021

  7. [7]

    Minimizing age of incorrect information over a channel with random delay,

    ——, “Minimizing age of incorrect information over a channel with random delay,”IEEE/ACM Transactions on Networking, 2024

    work page 2024

  8. [8]

    Age of incorrect information for remote estimation of a binary markov source,

    C. Kam, S. Kompella, and A. Ephremides, “Age of incorrect information for remote estimation of a binary markov source,” inProc. of IEEE Conference on Computer Communications Workshops, 2020

    work page 2020

  9. [9]

    Minimizing the age of incorrect information with continual belief learning,

    W. de Sombre, S. Dongare, A. Klein, and A. Ortiz, “Minimizing the age of incorrect information with continual belief learning,” inProc. of IEEE International Conference on Communications, 2025

    work page 2025

  10. [10]

    Minimizing the age of incorrect information for unknown markovian source,

    S. Kriouile and M. Assaad, “Minimizing the age of incorrect information for unknown markovian source,” 2022. [Online]. Available: https://arxiv.org/abs/2210.09681

    work page arXiv 2022

  11. [11]

    Semantics-empowered communications through the age of incorrect information,

    A. Maatouk, M. Assaad, and A. Ephremides, “Semantics-empowered communications through the age of incorrect information,” inProc. of IEEE International Conference on Communications, 2022

    work page 2022

  12. [12]

    A unified approach to learn transmission strategies using age-based metrics in point-to-point wireless communication,

    W. de Sombre, F. Marques, F. Pyttel, A. Ortiz, and A. Klein, “A unified approach to learn transmission strategies using age-based metrics in point-to-point wireless communication,”Proc. of IEEE Global Com- munications Conference, 2023

    work page 2023

  13. [13]

    Age of incorrect information minimization for semantic-empowered NOMA system in S-IoT,

    H. Hong, J. Jiao, T. Yang, Y . Wang, R. Lu, and Q. Zhang, “Age of incorrect information minimization for semantic-empowered NOMA system in S-IoT,”IEEE Transactions on Wireless Communications, 2024

    work page 2024

  14. [14]

    Age of incorrect information minimization for timely status updates in satellite-based IoT,

    Q. Yan, J. Jiao, T. Yang, L. An, Y . Wang, and Q. Zhang, “Age of incorrect information minimization for timely status updates in satellite-based IoT,” inProc. of IEEE/CIC International Conference on Communications in China, 2024

    work page 2024

  15. [15]

    Energy-age of incorrect information trade-off with timer-based sleep-wake scheduling,

    O. Guersoy and N. Akar, “Energy-age of incorrect information trade-off with timer-based sleep-wake scheduling,” 2024. [Online]. Available: http://dx.doi.org/10.36227/techrxiv.172503953.39448282/v1

    work page doi:10.36227/techrxiv.172503953.39448282/v1 2024

  16. [16]

    Age-optimal HARQ design for freshness-critical satellite-IoT systems,

    D. Li, S. Wu, Y . Wang, J. Jiao, and Q. Zhang, “Age-optimal HARQ design for freshness-critical satellite-IoT systems,”IEEE Internet of Things Journal, 2020

    work page 2020

  17. [17]

    A review of wireless and satellite-based m2m/iot services in support of smart grids,

    K. Sohraby, D. Minoli, B. Occhiogrossoet al., “A review of wireless and satellite-based m2m/iot services in support of smart grids,”Mobile Networks and Applications, 2018

    work page 2018

  18. [18]

    IoT enabled power quality monitoring in substations through LoRa LEO satellite communication,

    B. Ramachandran, R. M. Hussain, V . Subramanian, and J. V . Karuna- murthy, “IoT enabled power quality monitoring in substations through LoRa LEO satellite communication,” inProc. of 12th Electrical Power, Electronics, Communications, Controls and Informatics Seminar, 2024

    work page 2024

  19. [19]

    Real-time status: How often should one update?

    S. Kaul, R. Yates, and M. Gruteser, “Real-time status: How often should one update?” inProc. of IEEE Int. Conf. Computer Commun., 2012

    work page 2012

  20. [20]

    Scheduling policies for minimizing age of information in broadcast wireless networks,

    I. Kadota, A. Sinha, E. Uysal-Biyikoglu, R. Singh, and E. Modiano, “Scheduling policies for minimizing age of information in broadcast wireless networks,”IEEE/ACM Transactions on Networking, 2018

    work page 2018

  21. [21]

    The Age of Incorrect Information: A New Performance Metric for Status Updates,

    A. Maatouk, S. Kriouile, M. Assaad, and A. Ephremides, “The Age of Incorrect Information: A New Performance Metric for Status Updates,” IEEE/ACM Transactions on Networking, 2020

    work page 2020

  22. [22]

    Toward goal- oriented semantic communications: New metrics, framework, and open challenges,

    A. Li, S. Wu, S. Meng, R. Lu, S. Sun, and Q. Zhang, “Toward goal- oriented semantic communications: New metrics, framework, and open challenges,”IEEE Wireless Communications, 2024

    work page 2024

  23. [23]

    End-to-end latency evaluation of leo satellite iot systems,

    S. Heine, C. Hofmann, and A. Knopp, “End-to-end latency evaluation of leo satellite iot systems,” inProc. of 40th International Communications Satellite Systems Conference, 2023

    work page 2023

  24. [24]

    Downlink analysis and evaluation of multi-beam LEO satellite communication in shadowed rician channels,

    E. Kim, I. P. Roberts, and J. G. Andrews, “Downlink analysis and evaluation of multi-beam LEO satellite communication in shadowed rician channels,”IEEE Transactions on Vehicular Technology, 2024

    work page 2024

  25. [25]

    Simulation opti- mization: a review of algorithms and applications,

    S. Amaran, N. V . Sahinidis, B. Sharda, and S. J. Bury, “Simulation opti- mization: a review of algorithms and applications,”Annals of Operations Research, 2016

    work page 2016

  26. [26]

    Finite-state markov channel-a useful model for radio communication channels,

    H. S. Wang and N. Moayeri, “Finite-state markov channel-a useful model for radio communication channels,”IEEE Transactions on Ve- hicular Technology, 1995

    work page 1995

  27. [27]

    FIIT ESP32 kit hardware documentation,

    M. Pon ˇc´ak, “FIIT ESP32 kit hardware documentation,” https://github. com/999matej999/fiit-esp32-kit, accessed: Oct. 2025

    work page 2025

  28. [28]

    Power profiler kit II (PPK2) user guide,

    Nordic Semiconductor, “Power profiler kit II (PPK2) user guide,” https: //infocenter.nordicsemi.com/topic/ug ppk2/UG/ppk2/PPK2 intro.html, accessed: Nov. 2025

    work page 2025