pith. sign in

arxiv: 2407.08587 · v3 · submitted 2024-07-11 · 💻 cs.IT · cs.NI· math.IT

From AoI to QVAoI: Query-Based Semantics-Aware Scheduling for Energy-Harvesting IoT Systems

Erfan Delfani , Nikolaos Pappas This is my paper

Pith reviewed 2026-05-23 22:44 UTC · model grok-4.3

classification 💻 cs.IT cs.NImath.IT
keywords Query Version Age of InformationAge of InformationEnergy HarvestingIoT SystemsMarkov Decision ProcessSemantics-Aware SchedulingStatus UpdatesTransmission Policy
0
0 comments X

The pith

Optimizing for Query Version Age of Information yields fresher relevant updates or fewer transmissions than simpler age metrics in energy-harvesting IoT systems.

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

The paper establishes that a new metric combining query relevance with version age improves scheduling decisions for devices that harvest energy and send status updates. It formulates the transmission timing choice as a Markov Decision Process whose solution accounts for battery limits and energy arrivals while tracking how well updates answer destination queries. Comparisons of the resulting QVAoI-Optimal policy against policies based on QAoI, VAoI, AoI, and a greedy baseline show that it either supplies updates of higher freshness, relevance, and value with the same energy or meets the same quality targets with fewer transmissions. Closed-form expressions for average update rate and QVAoI under unit battery capacity supply analytical benchmarks. The approach matters because many IoT deployments must stretch harvested energy across updates that actually matter for queries rather than merely minimizing age.

Core claim

The central claim is that the QVAoI-Optimal policy demonstrates a significant performance improvement either by providing fresher, more relevant, and more valuable updates with the same energy arrivals or by reducing the number of transmissions in the system while maintaining the same level of freshness and information significance as the QAoI-Optimal and other policies.

What carries the argument

The Query Version Age of Information (QVAoI) metric, which augments age measures with query arrivals and version semantics and is minimized via a Markov Decision Process subject to energy harvesting constraints.

If this is right

  • All semantics-aware policies achieve better performance than the greedy policy.
  • The QVAoI-Optimal policy can maintain the same freshness and significance levels with fewer transmissions than QAoI-Optimal and other policies.
  • Closed-form expressions for average update rate and QVAoI provide analytical benchmarks for the unit-capacity battery case.
  • Semantics-aware policies deliver updates that are fresher, more relevant, and more valuable with fixed energy arrivals.

Where Pith is reading between the lines

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

  • The MDP formulation could be extended to include stochastic channel fading to check whether the reported gains persist outside idealized energy models.
  • Query-driven scheduling may generalize to multi-source or multi-destination IoT networks where different queries have distinct version requirements.
  • Adaptive versions of the policy could respond to observed changes in query statistics without recomputing the full MDP solution each time.

Load-bearing premise

The model assumes query arrivals and version semantics can be captured by a finite-state Markov Decision Process whose optimal policy remains effective under real channel and energy variability not detailed in the abstract.

What would settle it

A deployment test showing that the QVAoI-Optimal policy neither supplies higher-value updates nor reduces transmissions relative to the QAoI-Optimal policy under measured energy arrivals and channel conditions would disprove the performance advantage.

Figures

Figures reproduced from arXiv: 2407.08587 by Erfan Delfani, Nikolaos Pappas.

Figure 1
Figure 1. Figure 1: The considered system model. III. SYSTEM MODEL A. System Setup We consider a status update system, as depicted in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The resulting Markov chain with threshold [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The resulting Markov chain with threshold [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The resulting Markov chain with threshold [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Average QVAoI vs. pt. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.5 1 1.5 2 2.5 3 [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average VAoI vs. β. and ∆max = 19. Here, the semantics-aware policies also demonstrate superior performance compared to the greedy policy, and the QVAoI-optimal policy outperforms the QAoI￾optimal policy for lower values of pt. These results emphasize that incorporating semantics-aware metrics, leads to a more effective status updating policy con￾cerning the freshness and significance of information within… view at source ↗
Figure 9
Figure 9. Figure 9: Average QVAoI vs. β. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.2 0.4 0.6 0.8 1 1.2 1.4 [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
Figure 12
Figure 12. Figure 12: Average QVAoI vs. q. ps = 1, bmax = 15, and ∆max = 19. It can be seen that the average QVAoI increases with q for all three policies, with the QAoI-optimal policy demonstrating superior performance. The reason behind this increasing behavior is noteworthy. The first reason is that the query process r(t) emerges as a weight in the objective function of the MDP problem (6). As the probability of query arriv… view at source ↗
Figure 16
Figure 16. Figure 16: Average QVAoI and update rate for single-sized batt [PITH_FULL_IMAGE:figures/full_fig_p008_16.png] view at source ↗
Figure 14
Figure 14. Figure 14: Average VAoI for different policies vs. ps. 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 2.5 3 3.5 [PITH_FULL_IMAGE:figures/full_fig_p008_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Average QVAoI for different policies vs. [PITH_FULL_IMAGE:figures/full_fig_p008_15.png] view at source ↗
read the original abstract

In this work, we study the freshness and significance of information in an IoT status update system in which an Energy Harvesting (EH) device samples an information source and forwards update packets to a destination node via a direct channel. We introduce and optimize a semantics-aware metric, Query Version Age of Information (QVAoI), in the system along with other metrics: Query Age of Information (QAoI), Version Age of Information (VAoI), and Age of Information (AoI). We formulate the optimization problem as a Markov Decision Process to determine the optimal transmission policy at the device, which decides the time slots for transmitting updates, subject to the device's battery energy limitations and the energy arrivals. Furthermore, we derive closed-form expressions for the average update rate and the QVAoI for a unit-capacity battery, serving as analytical benchmarks. We compare the performance of QVAoI-Optimal, QAoI-Optimal, VoI-Optimal, and AoI-Optimal policies with a baseline greedy policy. All semantics-aware policies achieve better performance than the greedy policy. The QVAoI-Optimal policy, in particular, demonstrates a significant performance improvement either by providing fresher, more relevant, and more valuable updates with the same energy arrivals or by reducing the number of transmissions in the system while maintaining the same level of freshness and information significance as the QAoI-Optimal and other policies.

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 introduces the Query Version Age of Information (QVAoI) metric for semantics-aware status updates in an energy-harvesting IoT system. It formulates the transmission scheduling problem as a finite-state MDP whose optimal policy (QVAoI-Optimal) decides when to transmit subject to battery constraints and energy arrivals. Closed-form expressions are derived for the average update rate and QVAoI under unit-capacity battery. Simulations compare QVAoI-Optimal against QAoI-Optimal, VoI-Optimal, AoI-Optimal, and a greedy baseline, claiming that QVAoI-Optimal either delivers fresher/more valuable updates with the same energy or reduces transmissions while preserving freshness and significance.

Significance. If the MDP optimality and closed-forms are verified, the work supplies a new semantics-aware metric that jointly accounts for query relevance and version age, together with analytical benchmarks for the unit-battery case. The explicit comparison of four semantics-aware policies against greedy provides concrete evidence that incorporating query and version information can improve the freshness-significance-energy trade-off in EH IoT.

major comments (2)
  1. [Abstract] Abstract: the central performance claim that 'QVAoI-Optimal demonstrates a significant performance improvement' rests on the optimality of the finite-state MDP policy, yet no section demonstrates that this policy remains near-optimal when the channel is replaced by a non-i.i.d. fading process or energy arrivals deviate from the modeled distribution; the reported gains could therefore be an artifact of the discrete-state abstraction.
  2. [Abstract] Abstract: closed-form expressions are supplied only for unit-capacity battery, but the manuscript does not indicate how (or whether) the same expressions or the MDP policy extend to arbitrary battery capacities that are used in the general simulations; this limits the analytical support for the cross-policy comparison.
minor comments (1)
  1. [Abstract] The abstract states that 'all semantics-aware policies achieve better performance than the greedy policy' but does not reference the specific figures or tables that quantify the improvement; adding such pointers would improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below and will revise the manuscript to improve clarity on the scope of our results.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central performance claim that 'QVAoI-Optimal demonstrates a significant performance improvement' rests on the optimality of the finite-state MDP policy, yet no section demonstrates that this policy remains near-optimal when the channel is replaced by a non-i.i.d. fading process or energy arrivals deviate from the modeled distribution; the reported gains could therefore be an artifact of the discrete-state abstraction.

    Authors: The optimality of the QVAoI-Optimal policy and all reported performance gains are established exactly for the i.i.d. channel and energy-arrival model under which the finite-state MDP is formulated. The discrete-state MDP yields the precise optimal policy within this model rather than an approximation. We agree that the abstract should explicitly qualify the modeling assumptions to avoid any implication of robustness beyond the considered setting, and we will revise the abstract accordingly. revision: yes

  2. Referee: [Abstract] Abstract: closed-form expressions are supplied only for unit-capacity battery, but the manuscript does not indicate how (or whether) the same expressions or the MDP policy extend to arbitrary battery capacities that are used in the general simulations; this limits the analytical support for the cross-policy comparison.

    Authors: The closed-form expressions are derived only for the unit-capacity battery as analytical benchmarks for that special case. The underlying MDP formulation is defined for arbitrary finite battery capacities, and the optimal policies used in the general simulations are obtained directly from the MDP for each battery size. The cross-policy comparisons therefore rest on the MDP solutions rather than on the unit-battery closed forms. We will add a clarifying statement in the abstract and introduction to distinguish these two contributions. revision: yes

Circularity Check

0 steps flagged

No circularity; MDP optimization and closed-forms are independent of baselines

full rationale

The paper introduces QVAoI as a distinct metric, formulates a finite-state MDP over discrete battery, AoI, version age and query states, solves for the QVAoI-optimal policy, and supplies closed-form expressions only for the unit-capacity battery case as analytical benchmarks. These quantities are derived directly from the MDP transition structure and are not obtained by fitting parameters to the QAoI or VoI baselines. Policy comparisons are performed by evaluating each metric-specific optimal policy on the shared state space; no equation reduces a reported performance gain to a quantity defined by the same fitted inputs, and no load-bearing step relies on self-citation chains. The derivation chain is therefore self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that an MDP with states for battery and version accurately represents the joint dynamics of energy harvesting and query-driven value; no free parameters or invented physical entities are stated in the abstract.

axioms (1)
  • domain assumption The joint process of energy arrivals, battery state, and query-driven information value can be modeled as a finite-state Markov Decision Process.
    Invoked to formulate the transmission policy optimization problem.

pith-pipeline@v0.9.0 · 5804 in / 1200 out tokens · 26843 ms · 2026-05-23T22:44:58.291961+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

25 extracted references · 25 canonical work pages

  1. [1]

    Age of information: An introduction and survey,

    R. D. Y ates, Y . Sun, D. R. Brown, S. K. Kaul, E. Modiano, and S. Ulukus, “Age of information: An introduction and survey, ” IEEE Journal on Selected Areas in Commun. , vol. 39, no. 5, 2021. 9

    work page 2021

  2. [2]

    On the ro le of age of information in the internet of things,

    M. A. Abd-Elmagid, N. Pappas, and H. S. Dhillon, “On the ro le of age of information in the internet of things,” IEEE Commun. Magazine , vol. 57, no. 12, 2019

    work page 2019

  3. [3]

    Semantics-empowered comm unication for networked intelligent systems,

    M. Kountouris and N. Pappas, “Semantics-empowered comm unication for networked intelligent systems,” IEEE Commun. Magazine , vol. 59, no. 6, 2021

    work page 2021

  4. [4]

    Real-time status: Ho w often should one update?

    S. Kaul, R. Y ates, and M. Gruteser, “Real-time status: Ho w often should one update?” in 2012 Proceedings IEEE INFOCOM , 2012

    work page 2012

  5. [5]

    Closed-form anal ysis of non- linear age of information in status updates with an energy ha rvesting transmitter,

    X. Zheng, S. Zhou, Z. Jiang, and Z. Niu, “Closed-form anal ysis of non- linear age of information in status updates with an energy ha rvesting transmitter,” IEEE Trans. on Wireless Commun. , vol. 18, no. 8, 2019

    work page 2019

  6. [6]

    T he age of incorrect information: A new performance metric for status updates,

    A. Maatouk, S. Kriouile, M. Assaad, and A. Ephremides, “T he age of incorrect information: A new performance metric for status updates,” IEEE/ACM Trans. on Networking , vol. 28, no. 5, 2020

    work page 2020

  7. [7]

    Query age of information: Freshness in pul l-based communication,

    F. Chiariotti, J. Holm, A. E. Kalør, B. Soret, S. K. Jensen , T. B. Pedersen, and P . Popovski, “Query age of information: Freshness in pul l-based communication,” IEEE Trans. on Commun. , vol. 70, no. 3, 2022

    work page 2022

  8. [8]

    The age of gossip in networks,

    R. D. Y ates, “The age of gossip in networks,” in 2021 IEEE ISIT , 2021

    work page 2021

  9. [9]

    State-awar e timeliness in energy harvesting iot systems monitoring a markovian sourc e,

    E. Delfani, G. J. Stamatakis, and N. Pappas, “State-awar e timeliness in energy harvesting iot systems monitoring a markovian sourc e,” arXiv preprint arXiv:2405.03628, 2024

    work page arXiv 2024

  10. [10]

    A comprehensive survey on age of information in massive iot ne tworks,

    Q. Abbas, S. A. Hassan, H. K. Qureshi, K. Dev, and H. Jung, “A comprehensive survey on age of information in massive iot ne tworks,” Computer Commun. , vol. 197, 2023

    work page 2023

  11. [11]

    Pappas, M

    N. Pappas, M. A. Abd-Elmagid, B. Zhou, W. Saad, and H. S. D hillon, Age of Information: F oundations and Applications . Cambridge Uni- versity Press, 2022

    work page 2022

  12. [12]

    V ersion age o f information in clustered gossip networks,

    B. Buyukates, M. Bastopcu, and S. Ulukus, “V ersion age o f information in clustered gossip networks,” IEEE Journal on Selected Areas in Information Theory , vol. 3, no. 1, 2022

    work page 2022

  13. [13]

    Timely gossiping with file slic ing and network coding,

    P . Kaswan and S. Ulukus, “Timely gossiping with file slic ing and network coding,” in IEEE ISIT , 2022

    work page 2022

  14. [14]

    Susceptibility of age of gossip to timestomping,

    ——, “Susceptibility of age of gossip to timestomping,” in IEEE Information Theory W orkshop (ITW) , 2022

    work page 2022

  15. [15]

    Asuman: Age sense updating mult iple access in networks,

    P . Mitra and S. Ulukus, “Asuman: Age sense updating mult iple access in networks,” in 58th Annual Allerton Conference on Communication, Control, and Computing , 2022

    work page 2022

  16. [16]

    Age of gossip in ring networks i n the presence of jamming attacks,

    P . Kaswan and S. Ulukus, “Age of gossip in ring networks i n the presence of jamming attacks,” in 56th Asilomar Conference on Signals, Systems, and Computers , 2022

    work page 2022

  17. [17]

    Age-aware gossiping in network topologies,

    P . Mitra and S. Ulukus, “Age-aware gossiping in network topologies,” IEEE Trans. on Commun. , 2024

    work page 2024

  18. [18]

    V ersion age-optimal cached s tatus updates in a gossiping network with energy harvesting sensor,

    E. Delfani and N. Pappas, “V ersion age-optimal cached s tatus updates in a gossiping network with energy harvesting sensor,” in 2023 21st WiOpt Conference, 2023

    work page 2023

  19. [19]

    A learning based scheme for fair timeliness in sparse gossip networks,

    P . Mitra and S. Ulukus, “A learning based scheme for fair timeliness in sparse gossip networks,” arXiv preprint arXiv:2310.01396 , 2023

    work page arXiv 2023

  20. [20]

    Distribution of th e age of gossip in networks,

    M. A. Abd-Elmagid and H. S. Dhillon, “Distribution of th e age of gossip in networks,” Entropy, vol. 25, no. 2, 2023

    work page 2023

  21. [21]

    On-demand aoi minimization in resource-constrained cach e-enabled iot networks with energy harvesting sensors,

    M. Hatami, M. Leinonen, Z. Chen, N. Pappas, and M. Codrea nu, “On-demand aoi minimization in resource-constrained cach e-enabled iot networks with energy harvesting sensors,” IEEE Trans. on Commun. , vol. 70, no. 11, 2022

    work page 2022

  22. [22]

    P ull or wait: How to optimize query age of information,

    M. E. Ildiz, O. T. Y avascan, E. Uysal, and O. T. Kartal, “P ull or wait: How to optimize query age of information,” IEEE Journal on Selected Areas in Information Theory , 2023

    work page 2023

  23. [23]

    Implem entation and evaluation of age-aware downlink scheduling policies in pu sh-based and pull-based communication,

    T. K. O˘ guz, E. T. Ceran, E. Uysal, and T. Girici, “Implem entation and evaluation of age-aware downlink scheduling policies in pu sh-based and pull-based communication,” Entropy, vol. 24, no. 5, 2022

    work page 2022

  24. [24]

    E ffective communication: When to pull updates?

    P . Agheli, N. Pappas, P . Popovski, and M. Kountouris, “E ffective communication: When to pull updates?” in IEEE ICC , 2024

    work page 2024

  25. [25]

    D. P . Bertsekas, Dynamic Programming and Optimal Control, V ol. II , 3rd ed. Athena Scientific, 2007. APPENDIX A PROOF OF THEOREM 1 Proof. The Bellman equation at state s = (b, ∆, r) is given by: J ∗ +V (s) = min a∈{ 0,1} { ∑ s′∈ S P [ s′⏐ ⏐s, a ]( r∆′+V (s′) )    △ = Q(a) } (27) a∗ (s) = arg min a∈{ 0,1} Q(a) = { 0, DV (s) ≥ 0, 1, DV (s) < 0, (28) wh...

    work page 2007