Pith. sign in

REVIEW 2 major objections 4 minor 16 references

HAPS takes over coverage, cutting cellular power by 12.5%

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · glm-5.2

2026-07-08 17:20 UTC pith:YPVQ57J5

load-bearing objection Solid concept and public simulator, but the headline energy-savings number rests on undisclosed HAPS power parameters — worth a serious referee who pushes for a sensitivity analysis. the 2 major comments →

arxiv 2607.06072 v1 pith:YPVQ57J5 submitted 2026-07-07 cs.IT math.IT

HAPS as a Hypercell: Enabling Coverage and Capacity Carrier Shutdown in Cellular Networks

Matteo Bernab\`e , David L\'opez-P\'erez , Nicola Piovesan This is my paper
classification cs.IT math.IT
keywords HAPScarrier shutdownenergy efficiencynon-terrestrial networks5G6Gcellular networkspower consumption
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper proposes using a high-altitude platform station (HAPS) — an aerial 5G radio platform operating from the stratosphere — not for rural coverage extension but as an energy-saving tool for dense urban networks. The core idea is the HAPS-Hypercell: a wide-area non-terrestrial coverage layer that can assume the coverage role of multiple terrestrial macro-cells. In conventional cellular networks, coverage cells (high-power macro-cells) must stay continuously active to guarantee service, even during low-traffic hours, while only capacity cells can be shut down. By delegating the coverage function to the HAPS, both coverage and capacity terrestrial cells become eligible for shutdown. The authors formalize this concept, define two pairing architectures (non-hierarchical and hierarchical) that govern how terrestrial cells coordinate shutdown and reactivation with the HAPS, and develop a 3GPP-compliant system-level simulation to quantify the gains. Their results show up to 12.5% reduction in total network power consumption during low-traffic early-morning hours, with mean user rate reductions kept below 1.3% under conservative settings, demonstrating that the coverage function can be decoupled from terrestrial infrastructure without significant service degradation.

Core claim

The central finding is that a single HAPS platform, acting as a wide-area coverage layer over a dense urban network of 114 terrestrial cells, can enable the shutdown of both coverage-layer (4G) and capacity-layer (5G) macro-cells simultaneously. In 3GPP-compliant simulations, this yields up to 12.5% total network power savings during off-peak hours and approximately 10.5% on a 24-hour average under aggressive shutdown thresholds, while maintaining quality of service under conservative settings. The paper also discovers a structural trade-off between its two pairing architectures: the non-hierarchical design (all cells paired directly with HAPS) achieves the largest energy savings but suffers

What carries the argument

HAPS-Hypercell (a logical coverage entity supported by a HAPS platform that ensures service continuity independently of terrestrial cells); two pairing architectures — HAPS-NH (non-hierarchical: all terrestrial cells paired directly with HAPS as capacity nodes) and HAPS-H (hierarchical: capacity cells paired with coverage cells, coverage cells paired with HAPS); distributed carrier shutdown algorithm operating over discrete decision intervals with shutdown and wake-up thresholds; 3GPP-compliant system model with UMa channel models, mMIMO precoding, and a multi-carrier power consumption model.

Load-bearing premise

The power consumption model for the HAPS and terrestrial cells uses numerical parameters based on commercial products that are omitted for confidentiality, meaning the headline energy savings are contingent on specific but undisclosed power figures. If the HAPS platform itself consumes substantial power, the net savings could be significantly smaller or negative.

What would settle it

If the HAPS platform's own power consumption (P_BBU, P_0, P_TRX, P_PA in Eq. 12) is high enough that it exceeds the combined savings from shutting down multiple terrestrial coverage and capacity cells, the net network power reduction would vanish or reverse. The 12.5% savings figure is directly falsifiable once the confidential power parameters are disclosed.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If a HAPS can reliably hold the coverage function for a dense urban area, the traditional always-on macro-cell layer becomes optional rather than mandatory, fundamentally changing how operators dimension and power their networks during off-peak hours.
  • The pairing architecture trade-off suggests that future HAPS-integrated networks will need adaptive control logic that switches between hierarchical and non-hierarchical modes depending on traffic load, rather than committing to one architecture permanently.
  • The finding that local PRB-load-based decisions are insufficient for full savings implies that richer information exchange protocols between the HAPS and terrestrial cells will be needed to realize the full potential of this approach in operational networks.
  • The energy savings of 10.5% on a 24-hour average, if achievable at scale, would translate to roughly 30-35 TWh of annual electricity reduction across global mobile networks, a material contribution to the ICT sector's sustainability targets.

Where Pith is reading between the lines

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

  • The net energy savings are entirely contingent on the HAPS platform's own power consumption being low enough relative to the terrestrial cells it replaces. The paper uses commercial power model parameters that are omitted for confidentiality, making the headline savings figure unverifiable without that data. If the HAPS power figure is high, the net savings could vanish or reverse.
  • The single-sector HAPS covering the entire 57-cell area creates a single point of failure for coverage. The paper does not address what happens if the HAPS platform experiences an outage, which would leave the network with no coverage layer at all.
  • The 500m inter-site distance and 25m cell height represent a specific dense urban scenario. The savings would likely differ substantially in suburban or rural deployments where terrestrial cell density is lower and the relative cost of maintaining coverage cells is different.
  • The convergence within tens of iterations claimed for the distributed algorithm is stated for the analyzed scenarios; convergence behavior under rapid traffic spikes or correlated failure modes is not characterized and could be a practical deployment concern.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. This paper proposes using a high-altitude platform station (HAPS) as a wide-area non-terrestrial coverage layer (a 'HAPS-Hypercell') to enable the joint shutdown of both capacity and coverage terrestrial macro-cells during low-traffic periods. The authors formalize two pairing architectures (non-hierarchical and hierarchical) and a distributed carrier shutdown (CS) algorithm. The system is evaluated using a 3GPP-compliant system-level simulator (publicly available) over 24-hour traffic cycles. Results indicate power savings of up to 12.5% relative to a terrestrial network without CS, while maintaining QoS under conservative thresholds.

Significance. The core concept of decoupling the coverage role from terrestrial infrastructure and delegating it to a HAPS to enable coverage-cell shutdown is a novel and practically motivated contribution to green networking. The provision of a publicly available, high-fidelity system-level simulator (Giulia) is a significant strength that enhances reproducibility. The exploration of trade-offs between non-hierarchical and hierarchical pairing architectures under varying traffic loads provides useful design insights for next-generation CS operations.

major comments (2)
  1. Section III-7, Eq. (12): The central energy savings claim (up to 12.5% power reduction) is entirely contingent on the net power balance between the terrestrial cells deactivated and the HAPS power consumed. However, the numerical values for the power consumption parameters are stated to be 'omitted for confidentiality.' While terrestrial parameters can be cross-referenced to [15], the HAPS-specific parameters (e.g., P_BBU, P_0, P_TRX, D_TRX, M_av_TRX, D_PA, M_ac_PA, η for the HAPS radio unit) have no external reference and cannot be verified. Without these values, the headline number is conditional on an unverifiable assumption. A sensitivity analysis over plausible HAPS power parameters (e.g., drawn from 3GPP TR 38.811 or HAPS literature such as [5]) is necessary to establish the robustness of the claimed savings.
  2. Section III-7, Eq. (12): It is unclear whether the HAPS power consumption shown in Figure 2 includes only the radio unit modeled by Eq. (12) or also accounts for platform-level power (propulsion, station-keeping). If the HAPS is solar-powered, platform power may legitimately be excluded from grid consumption, but this assumption is not stated. The manuscript should explicitly clarify the scope of the HAPS power model.
minor comments (4)
  1. Section III-1: The HAPS is described as a 'single-sector' system. Given that it covers 19 sites / 57 cells per layer, clarifying the HAPS antenna array configuration (number of elements, transceivers) in the context of Eq. (12) would help the reader understand the HAPS power scaling.
  2. Table II: The threshold values are defined as θ_shtdn/δ_c and θ_wkup/δ_b. The notation in the table header is slightly ambiguous; ensuring it matches the definitions in Eq. (5) and the surrounding text would improve clarity.
  3. Figure 2: The y-axis ranges differ between the low traffic (55-65 kW) and high traffic (60-70 kW) subfigures. While this is common, a brief note in the caption or consistent axes would aid direct visual comparison.
  4. Section IV-A: The text states that HAPS-NH reaches '10.5% on a 24-hour average relative to TN No CS.' It would be useful to also report the 24-hour average for the conservative and balanced profiles to give a fuller picture of the trade-off space.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and the constructive feedback. Both major comments are well-taken and address legitimate concerns about the verifiability and scope of our power model. We address each below.

read point-by-point responses

  1. Referee: Section III-7, Eq. (12): The central energy savings claim (up to 12.5% power reduction) is entirely contingent on the net power balance between the terrestrial cells deactivated and the HAPS power consumed. However, the numerical values for the power consumption parameters are stated to be 'omitted for confidentiality.' While terrestrial parameters can be cross-referenced to [15], the HAPS-specific parameters (e.g., P_BBU, P_0, P_TRX, D_TRX, M_av_TRX, D_PA, M_ac_PA, η for the HAPS radio unit) have no external reference and cannot be verified. Without these values, the headline number is conditional on an unverifiable assumption. A sensitivity analysis over plausible HAPS power parameters (e.g., drawn from 3GPP TR 38.811 or HAPS literature such as [5]) is necessary to establish the robustness of the claimed savings.

    Authors: The referee is correct that the headline savings are contingent on the HAPS power parameters, and that omitting these values without any compensating analysis limits verifiability. We accept this point and will add a sensitivity analysis in the revised manuscript. Specifically, we will sweep the HAPS radio-unit power parameters over plausible ranges derived from 3GPP TR 38.811 and the HAPS energy literature (e.g., [5], [6]) and report the resulting range of network power savings. This will make explicit the conditions under which the 12.5% figure holds and the break-even point at which HAPS power consumption would negate the terrestrial savings. We will also provide a table summarizing the parameter ranges used. We note that the terrestrial parameters, which dominate the power balance in our dense urban scenario (114 terrestrial cells vs. a single HAPS sector), are cross-referenced to [15] and are the primary driver of the savings magnitude; the HAPS contribution is comparatively small. Nevertheless, we agree that this must be demonstrated quantitatively rather than asserted. revision: yes

  2. Referee: Section III-7, Eq. (12): It is unclear whether the HAPS power consumption shown in Figure 2 includes only the radio unit modeled by Eq. (12) or also accounts for platform-level power (propulsion, station-keeping). If the HAPS is solar-powered, platform power may legitimately be excluded from grid consumption, but this assumption is not stated. The manuscript should explicitly clarify the scope of the HAPS power model.

    Authors: The referee raises a valid point that we failed to state explicitly. The HAPS power consumption in our model, as captured by Eq. (12) and reflected in Figure 2, includes only the radio unit (baseband, transceiver, power amplifier, and radiated power). It does not include platform-level power such as propulsion or station-keeping. This is a deliberate modeling choice grounded in the assumption, common in the HAPS literature, that the platform is solar-powered and that propulsion energy is sourced independently of the grid-connected RAN infrastructure whose consumption we aim to minimize. However, we agree that this assumption should have been stated explicitly. In the revised manuscript, we will add a clarifying paragraph in Section III-7 explaining the scope of the HAPS power model, the solar-power assumption, and the rationale for excluding platform-level power from the network energy balance. We will also acknowledge this as a limitation, since a grid-powered or battery-constrained HAPS would alter the net energy calculus. revision: yes

Circularity Check

0 steps flagged

No significant circularity; self-citations are for prior concepts and models, but the central energy-savings claim is a measured simulation outcome, not a definitional or fitted result.

full rationale

The paper's central claim—up to 12.5% network power reduction from HAPS-enabled joint coverage and capacity shutdown—is a measured outcome of 3GPP-compliant system-level simulations (Figure 2), not a quantity derived from a formula that embeds the answer. The derivation chain is: (1) Eq. 1 defines the Hypercell coverage condition (a design specification, not a prediction); (2) Eqs. 2-4 define pairing architectures (design choices); (3) Eq. 5 defines the shutdown/wake-up rule using thresholds from Table II (input parameters, not fitted-then-predicted); (4) Eqs. 6-13 are standard 3GPP channel, SINR, rate, and power models; (5) the energy savings are computed by the simulator as terrestrial-power-saved minus HAPS-power-added. Two self-citations appear: [6] (Song, López-Pérez, Meo, Piovesan, Renga) for the NTN Hypercell concept, and [15] (Piovesan, López-Pérez, et al.) for the power consumption model (Eq. 12). Neither is load-bearing in a circular sense: [6] provides the conceptual precursor that this paper extends with new pairing architectures, CS algorithms, and full system-level evaluation; [15] provides a power model that is an input to the simulation, not the output being claimed. The thresholds in Table II are configuration parameters defining CS profiles, not parameters fitted to a subset of data and then 'predicted' on related data. The undisclosed HAPS power parameters (Section III-7, 'omitted for confidentiality') raise a verifiability and correctness concern, but this is not circularity—the power model is an input and the savings are an output. No step in the chain reduces to its own inputs by construction.

Axiom & Free-Parameter Ledger

3 free parameters · 3 axioms · 1 invented entities

The ledger captures the key thresholds that define the CS profiles, the undisclosed power model parameters that gate the central claim, and the standard channel model assumptions underlying the simulation.

free parameters (3)
  • θ_shtdn (shutdown threshold) = 0.10, 0.30, 0.50 (Conservative, Balanced, Aggressive)
    Hand-set thresholds defining when cells can deactivate; directly control the aggressiveness of energy savings.
  • θ_wkup (wake-up threshold) = 0.60, 0.70, 0.90 (Conservative, Balanced, Aggressive)
    Hand-set thresholds defining when deactivated cells are reactivated; control QoS trade-offs.
  • HAPS power consumption parameters = Omitted for confidentiality
    Numerical values for P_BBU, P_0, P_BB, D_TRX, D_PA in Eq. 12 are based on commercial products but not disclosed, making the net energy savings unverifiable.
axioms (3)
  • domain assumption 3GPP TR38.901 and TR38.811 channel models accurately represent the terrestrial and NTN propagation environments.
    The entire link-level and system-level performance depends on these standard channel models being valid for the HAPS-to-urban scenario.
  • domain assumption The distributed CS algorithm converges to a stable state within 'tens of iterations' (Section II-D).
    No formal proof of convergence is provided; stability is asserted empirically.
  • domain assumption The HAPS has sufficient capacity to absorb traffic from deactivated terrestrial cells without becoming a bottleneck in low-to-medium traffic.
    The energy savings depend on the HAPS not saturating; the paper acknowledges this limitation under high traffic but assumes it holds in the low-traffic regime where savings are reported.
invented entities (1)
  • HAPS-Hypercell independent evidence
    purpose: A logical coverage entity supported by a wide-area NTN platform that assumes the coverage role of terrestrial macro-cells.
    The concept is evaluated through 3GPP-compliant simulations showing measurable energy savings, providing a falsifiable handle on its performance.

pith-pipeline@v1.1.0-glm · 13007 in / 2287 out tokens · 380133 ms · 2026-07-08T17:20:03.503528+00:00 · methodology

0 comments
read the original abstract

Energy consumption remains a dominant operational challenge for current and future cellular systems, especially in dense urban deployments. This paper investigates a novel role for non terrestrial network (NTN) high-altitude platform station (HAPS) as an enabler of energy-efficient operation rather than only coverage extension. We define the HAPS-Hypercell as a wide-area non-terrestrial layer that can assume the coverage role of multiple terrestrial macro-cells, enabling, for the first time, the shutdown of both capacity and coverage macro-cells. We develop a comprehensive third generation partnership project (3GPP)-compliant system model, along with two HAPS-Hypercell pairing architectures that capture the interplay among multiple layers, realistic channel conditions, and distributed carrier shutdown (CS) mechanisms. Our results show that the HAPS-Hypercell can effectively reduce overall network power consumption. We then identify key limitations of a straightforward HAPS integration, laying the groundwork for future optimization and providing key insights for next-generation CS operations.

Figures

Figures reproduced from arXiv: 2607.06072 by David L\'opez-P\'erez, Matteo Bernab\`e, Nicola Piovesan.

Figure 1
Figure 1. Figure 1: Illustration of the proposed HAPS-Hypercell pairing architectures. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Average network power consumption over 24-hours under different CS profiles and traffic demand. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Resulting UE achievable rate distribution under different CS profiles and traffic demand. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

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

16 extracted references · 16 canonical work pages

  1. [1]

    Mobile Net Zero: State of the Industry on Climate Action 2024,

    GSMA, “Mobile Net Zero: State of the Industry on Climate Action 2024,” GSMA, Tech. Rep., 2024

    work page 2024

  2. [2]

    Data Centres and Data Transmission Networks,

    IEA, “Data Centres and Data Transmission Networks,” International Energy Agency (IEA), Tech. Rep., 2023

    work page 2023

  3. [3]

    Assessing ICT global emissions footprint: Trends to 2040 & recommendations,

    L. Belkhir and A. Elmeligi, “Assessing ICT global emissions footprint: Trends to 2040 & recommendations,”Journal of Cleaner Production, vol. 177, pp. 448–463, Sep. 2018

    work page 2040

  4. [4]

    Data-Driven Energy Efficiency Modeling in Large-Scale Networks: An Expert Knowledge and ML-Based Approach,

    D. L ´opez-P´erez, A. De Domenico, N. Piovesan, and M. Debbah, “Data-Driven Energy Efficiency Modeling in Large-Scale Networks: An Expert Knowledge and ML-Based Approach,”IEEE Trans. Mach. Learn. Commun. Netw., vol. 2, pp. 780–804, Jun. 2024

    work page 2024

  5. [5]

    Can High Altitude Platform Stations Make 6G Sustainable?

    D. Renga and M. Meo, “Can High Altitude Platform Stations Make 6G Sustainable?”IEEE Commun. Mag., vol. 60, no. 9, May 2022

    work page 2022

  6. [6]

    High altitude platform stations: the new network energy efficiency enabler in the 6G era,

    T. Song, D. Lopez, M. Meo, N. Piovesan, and D. Renga, “High altitude platform stations: the new network energy efficiency enabler in the 6G era,” inProc. IEEE Wireless Commun. Networking Conference (WCNC), Apr. 2024, pp. 1–6

    work page 2024

  7. [7]

    Haps-enabled sustainability provision in cellular networks through cell-switching,

    G. B. Koc ¸, B. C ¸ ilo˘glu, M. ¨Ozt¨urk, and H. Yanikomeroglu, “Haps-enabled sustainability provision in cellular networks through cell-switching,” in Proc. IEEE Int. Black Sea Conf. on Commun. and Netw. (BlackSeaCom), Jul. 2023, pp. 294–299

    work page 2023

  8. [8]

    Energy Sustainability in Dense Radio Access Networks via High Altitude Platform Stations,

    M. Salamatmoghadasi, A. Mehrabian, and H. Yanikomeroglus, “Energy Sustainability in Dense Radio Access Networks via High Altitude Platform Stations,”IEEE Net. Letters, vol. 6, no. 1, Mar. 2024

    work page 2024

  9. [9]

    Cell Switching in HAPS-Aided Networking: How the Obscurity of Traffic Loads Affects the Decision,

    B. C ¸ ilo˘glu, G. B. Koc ¸, M. Ozturk, and H. Yanikomeroglu, “Cell Switching in HAPS-Aided Networking: How the Obscurity of Traffic Loads Affects the Decision,”IEEE Trans. Veh. Technol., vol. 73, no. 11, pp. 17 782–17 787, Jul. 2024

    work page 2024

  10. [10]

    Sustainable Vertical Heterogeneous Networks: A Cell Switching Approach With High Altitude Platform Station,

    M. Salamatmoghadasi, A. Mehrabian, H. Yanikomeroglu, and G. Kad- doum, “Sustainable Vertical Heterogeneous Networks: A Cell Switching Approach With High Altitude Platform Station,”IEEE Trans. Green. Comm. and Net., vol. 10, pp. 1951–1966, Jan. 2026

    work page 1951

  11. [11]

    Study on channel model for frequencies from 0.5 to 100 GHz,

    3GPP TR38.901, “Study on channel model for frequencies from 0.5 to 100 GHz,” Mar. 2017, v.14.0

    work page 2017

  12. [12]

    Study on New Radio (NR) to support non-terrestrial networks,

    3GPP TR38.811, “Study on New Radio (NR) to support non-terrestrial networks,” Sep. 2020, v.15.4

    work page 2020

  13. [13]

    Energy efficiency analysis of the reference systems, areas of improvements and target breakdown,

    W. M. Wajda, G. Auer, P. Skillermark, and Y . Jading, “Energy efficiency analysis of the reference systems, areas of improvements and target breakdown,” EARTH Project (INFSO-ICT-247733), European Commis- sion, Deliverable D2.3, 2012

    work page 2012

  14. [14]

    Dahlman, S

    E. Dahlman, S. Parkvall, and J. Skold,5G NR: The Next Generation Wireless Access Technology, 2nd ed. USA: Academic Press, Inc., 2018

    work page 2018

  15. [15]

    Machine learning and analytical power consumption models for 5G base stations,

    N. Piovesan, D. L ´opez-P´erez, A. De Domenico, X. Geng, H. Bao, and M. Debbah, “Machine learning and analytical power consumption models for 5G base stations,”IEEE Commun. Mag., vol. 60, no. 10, pp. 56–62, Oct. 2022

    work page 2022

  16. [16]

    Network energy efficiency phase 2,

    NGMN Alliance, “Network energy efficiency phase 2,” Next Generation Mobile Networks Alliance, White Paper v1.0, Oct. 2023

    work page 2023