ORCHID: Fairness-Aware Orchestration in Mission-Critical Air-Ground Integrated Networks

Chi Jai Choy; Chuan-Chi Lai

REVIEW 4 major objections 6 minor 34 references

ORCHID: Fairness-Aware Orchestration in Mission-Critical Air-Ground Integrated Networks

T0 review · 4 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read Fairness-first drone orchestration can also be the most energy-efficient choice, reversing the usual efficiency-equity trade-off.

desk verdict Plausible two-stage MARL framework with a promising fairness-efficiency finding, but the promised EBG theory is missing, the R&F trigger looks fitted, and the PF comparison is under-specified. read the letter →

arxiv 2602.09994 v2 pith:QYJ3F7X2 submitted 2026-02-10 cs.NI

Chuan-Chi Lai , Chi Jai Choy This is my paper · ORCID

classification cs.NI

keywords UAVorchestrationair-groundintegratednetworksmulti-agentreinforcementlearningmax-minfairnessenergyefficiencyMAPPOReset-and-FinetuneThomasclusterprocess

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that, in mission-critical air-ground networks where users cluster in hotspots, orchestrating drones for max-min fairness instead of proportional fairness does not sacrifice energy efficiency—it improves it. The core is a two-stage learning framework, ORCHID: a GBS-aware K-Means++ initialization that drops drones near demand, and a Reset-and-Finetune mechanism inside MAPPO that resets the optimizer and shrinks the learning rate when fairness plateaus, suppressing the policy degradation that plagues multi-agent deep RL. The paper claims ORCHID-MMF beats ORCHID-PF by 6.8% in normalized energy efficiency while slightly increasing fairness, and that ORCHID overall dominates MADDPG and static heuristics on both axes. If true, the standard fairness-efficiency trade-off in UAV orchestration is reversed for hotspot-shaped missions, and a cheap stability trick makes the cooperation robust.

What carries the argument

Named mechanism: Reset-and-Finetune (R&F). A plateau detector watches the sliding-window average of the global Jain Fairness Index; when relative improvement drops below a threshold, it clears the Adam optimizer's momentum/variance states and steps down both learning rates by a factor 0.1, suppressing gradient variance quadratically. This lets MAPPO switch from exploration to fine-tuning and is what the paper credits for breaking through local optima. Phase I, GBS-aware K-Means++ clustering (keeping the N clusters farthest from the macro base station) provides the warm-start initialization that makes the fine-tuning stable.

What would settle it

Run ORCHID-MMF and ORCHID-PF under identical, published reward weights and penalty coefficients, with the R&F plateau detector (Equation 29) left fully adaptive and no fixed episode-500 trigger; if MMF no longer exceeds PF in normalized energy efficiency by roughly 6.8% across multiple seeds, the claimed synergy collapses.

Watch

Extended reading notes

Core claim

The central claim is that a Max-Min Fairness objective, realized by putting Jain's Fairness Index into the reward, makes the UAV fleet self-organize into a more dispersed, load-balanced topology that is also more energy-efficient than the Proportional Fairness objective (sum of log rates), which converges to greedy local equilibria. This is supported by simulations showing ORCHID-MMF achieves 6.8% higher Normalized Energy Efficiency (NEE) than ORCHID-PF (2.20 vs 2.06 relative units) and slightly higher user-rate JFI (0.403 vs 0.388), and that the full ORCHID framework—with its structured initialization and reset mechanism—sits in the upper-right (high fairness, high energy) region relative t

Load-bearing premise

The paper's conclusions rest on a single simulation configuration in which the reward weights, penalty coefficients, stability threshold, and the R&F trigger location are fixed and unreported, so the efficiency-fairness synergy and stability gain could be artifacts of that tuning rather than properties of the framework.

Editorial extensions

If this is right

If correct, mission-critical networks can enforce a no-user-left-behind policy without an energy penalty; maximum fairness and maximum energy efficiency become compatible in clustered scenarios.
A 6.8% NEE improvement over PF means drone fleets can serve edge users longer on the same battery budget, extending mission duration in disaster response.
ORCHID's convergence behavior suggests that MARL-based orchestration can be deployed with reliable late-stage stability, addressing the policy-degradation problem that limits practical adoption.
The Pareto-dominance over MADDPG and static heuristics supports the two-stage coarse-to-fine design as an effective practical template for UAV orchestration.
The theoretical account predicts MMF yields a more dispersed and load-balanced topology; this spatial property is what carries the efficiency gain.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A direct test: compute the spatial dispersion (e.g., mean pairwise UAV distance or coverage-area union) of the final policies under MMF vs PF; the paper claims but does not directly plot this topology difference, and it would isolate the mechanism behind the 6.8%.
The R&F trigger may be mimicking an adaptive learning-rate schedule; comparing R&F against simple cosine annealing or scheduled LR decay in the same MAPPO setup would show whether the optimizer-state reset is the active ingredient or just the decay.
The MMF-over-PF energy gain likely depends on the Thomas-cluster scatter; under near-uniform user distribution the greedy equilibrium of PF may not be harmful, so the synergy may not generalize—a parameter sweep over cluster intensity and scatter would map the regime.
Because the reward weights and penalty coefficients are not reported, an independent reimplementation cannot yet reproduce the numbers; publishing those values (or running a sensitivity ablation) would turn the claimed synergy into a repeatable finding.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Colleague,

The paper is a decent engineering draft with a real empirical question, but it currently overclaims both the theory and the evidence. The core idea — GBS-aware K-Means++ initialization followed by a MAPPO variant with optimizer reset and learning-rate decay — is coherent and worth testing, and the ablation study does show that removing either component hurts performance. The finding that an MMF objective can beat PF on energy efficiency in a clustered user topology is genuinely counterintuitive and, if it holds, would be a useful result for UAV orchestration.

That said, there are several load-bearing soft spots. Most obvious is the abstract-body mismatch: the introduction and abstract promise a theoretical analysis based on an Egalitarian Bargaining Game, but no such formulation appears anywhere in Sections III–VI. There is no theorem, no game-theoretic equilibrium analysis. That claim should be removed or actually delivered.

Second is the R&F trigger. Eq. (29) defines an adaptive plateau detector with a threshold ε_tol, but every convergence figure shows a vertical line at episode 500, and ε_tol is never reported. If the trigger was set by inspecting the training curves rather than by the algorithm's own rule, then the post-500 'leap' is just a scheduled learning-rate decay, and the paper's main methodological novelty is not demonstrated. The variance argument in Eq. (32) is also just scaling a learning rate — it is not a stability theorem.

Third, the PF comparison is under-specified. PF is described as 'maximize the sum of logarithmic data rates,' but the reward function (20) is a weighted sum of four terms plus penalties. It is unclear whether PF shares the same weights and only swaps the fairness term, or whether it omits the EE term. If the latter, the 6.8% NEE advantage is partly built into the comparison. The paper also never reports w1–w5, the penalty coefficients, the initial altitude, or the coverage threshold, so the results are not reproducible. And the statistical base is thin — five runs, no error bars on the scatter plot, and no significance test. There is also an internal inconsistency: Section V-F reports NEE values around 2.2×10^6, but NEE is defined as normalized to the random baseline, so it should be a dimensionless quantity around 2.4.

None of these issues kill the underlying idea. The framework is plausible and the ablation is informative. But the current evidence is insufficient to support the claims as written. The paper deserves a serious referee, mainly because the central question is worth asking and the authors have assembled a reasonable system model. If the authors can fix the framing, report the missing configuration, clarify the PF implementation, and add proper statistics — ideally with a code release — it could become a solid contribution. As it stands, it should not be accepted without major revision.

If you're running a reading group on fairness in MARL, this is a decent discussion piece — the 6.8% claim is a nice trap to unpack. I wouldn't cite it yet.

Best.

Referee Report

4 major / 6 minor

Summary. The paper proposes ORCHID, a two-stage MAPPO-based framework for multi-UAV orchestration in mission-critical air-ground integrated networks. Stage I initializes UAV positions with GBS-aware K-Means++ clustering; Stage II applies MAPPO with a Reset-and-Finetune (R&F) mechanism that resets Adam optimizer states and decays learning rates upon detecting a JFI plateau. The reward uses weighted coverage, energy-efficiency, load-fairness, and rate-fairness terms with penalties. The paper claims that the MMF objective yields both higher fairness and higher normalized energy efficiency than PF (6.8% NEE gain), that R&F stabilizes training, and that ORCHID is Pareto-dominant over MADDPG and static heuristics. The abstract and introduction also promise an Egalitarian Bargaining Game (EBG) theoretical analysis, but no EBG formulation, theorem, or fairness-efficiency proof appears in Sections III–VI.

Significance. If the empirical claims held, the paper would present a practically useful framework and a counter-intuitive result: MMF improving energy efficiency over PF in clustered UAV deployments, together with a lightweight stabilizer (optimizer reset plus learning-rate decay) that reverses late-training policy degradation. The ablation study in Fig. 6 and the TCP-based scenario are appropriate methodological choices, and the complexity analysis in Section IV-E is useful. However, the significance is currently conditional on reproducibility and on the validity of the MMF-vs-PF comparison. The absence of the advertised EBG analysis, the lack of reported reward/penalty weights, and the fixed R&F trigger at episode 500 all undermine the central claims as written. The paper also ships no code or pseudo-code details that would let a reader verify the experiments independently.

major comments (4)

[§V-A, §V-F, Eq. (20), Eq. (25)] The central MMF-vs-PF claim (6.8% NEE gain) is not verifiable because the reward weights w1–w5 and penalty coefficients ωc, ωb, ωbh are never reported, and the PF baseline is described only as 'maximize the sum of logarithmic data rates.' If ORCHID-PF uses a different reward (e.g., no EE term or different weights), the 6.8% NEE advantage is partly built into the comparison. Please report the exact reward/penalty values for both MMF and PF, and confirm that both variants use identical weight vectors except for the fairness objective.
[§IV-D, Eq. (29), Figs. 3/4/6] The R&F mechanism is introduced as an adaptive plateau detector, but all convergence figures show the trigger fixed at episode 500. The stability threshold ε_tol and window size W are not reported beyond W=50 in Table I, and no run-to-run variation in the trigger episode is shown. If the trigger was hand-set to episode 500 after inspecting training curves, the post-500 'leap' is a scheduled learning-rate decay rather than evidence of adaptive stabilization. Please report ε_tol, the distribution of trigger episodes across runs, and a version where the trigger is selected by Eq. (29) online.
[§IV-D, Eq. (32); abstract and §I] The advertised theoretical contribution—an EBG-based analysis of fairness and energy efficiency—does not appear anywhere in the body. Eq. (32) is a standard scaling property of stochastic gradient updates: scaling the learning rate by κ scales the parameter-update variance by κ², assuming gradient variance is unchanged. This does not explain why MMF yields a more dispersed, load-balanced topology, nor why MMF beats PF in energy efficiency. Please either supply the promised EBG formulation and a theorem connecting its solution to Eq. (32), or remove the EBG/theoretical-explanation claims from the abstract and introduction.
[§V-F, Eq. (35), Fig. 7] The reported NEE values (≈2.20×10^6 for MMF vs ≈2.06×10^6 for PF) are inconsistent with Eq. (35), where NEE is defined as a unitless ratio normalized by EE_random. A value of 2.20×10^6 suggests that either the normalization was not applied as stated or the vertical axis in Fig. 7(b) is raw EE. Moreover, the text calls the 3.86% JFI gain 'statistically significant' without reporting a significance test, confidence intervals, or the number of evaluation episodes beyond 5 runs. Please clarify the NEE units/definition and provide a proper statistical comparison.

minor comments (6)

[§II-C/§I] The EBG claim in the abstract and contribution list is never revisited. If the analysis is deferred or dropped, the text should be revised for consistency.
[Eq. (9) vs Eq. (26)] The symbol γ is used both for the terrestrial path-loss exponent (Eq. 9) and for the discount factor (Eq. 26). Please use distinct notations.
[§V-A] Table I omits several parameters that affect the results: the TCP parameters λp and σ_scatter and mean users per cluster, the coverage threshold Γ_req, initial altitude h_init, and the R&F stability threshold ε_tol. Please add them.
[§V-B, Eq. (33)] Coverage is defined using SINR in Eq. (33), but the system model in Eq. (10) defines only SNR. Please align the terminology.
[§V-C and Fig. 3] The text states that 'confidence interval significantly narrows' post-500, but only ±1 standard deviation over 5 runs is shown; no numerical variance reduction is reported. A quantitative before/after variance statistic would support the claim.
[General] There are several typos and formatting issues (e.g., 'Jain’s' capitalization, 'UA Vs' spacing, unreadable math glyphs in the PDF). A careful copyedit is needed.

Circularity Check

1 steps flagged · score 3.0 of 10

Only the R&F 'theoretical validation' reduces to the mechanism itself; the MMF-vs-PF empirical comparison is externally benchmarked and not circular, while the promised EBG analysis is absent and key hyperparameters are unreported.

other [Section IV-D (Reset-and-Finetune), Eq. (32)]
"Mathematically, by scaling η by κ, the variance of the parameter updates is suppressed quadratically: Var(Δθ′)≈ κ²·Var(Δθ) = 0.01·Var(Δθ). This order-of-magnitude reduction acts as a numerical damper, effectively locking the UAV fleet into a low-variance equilibrium, thereby validating the robustness of the R&F mechanism."

The R&F mechanism is defined by Eqs. (30)-(31): reset Adam moments and multiply η by κ=0.1 after e*. Eq. (32) then states Var(κ·Δθ)=κ²Var(Δθ), an algebraic identity for any scaled gradient step, and labels it 'theoretically' validating a 'low-variance equilibrium.' The claimed explanation is therefore a restatement of the chosen learning-rate decay, not a derivation from the system model or the MMF/PF comparison. It does not infect the empirical Section V benchmarks, which are external, so the circularity is confined to this explanatory equation.

full rationale

The central empirical claims -- ORCHID's Pareto dominance, 6.8% NEE gain of MMF over PF, and 3.86% JFI gain -- are supported by simulations against Static Random, Static K-Means++, and MADDPG baselines; those comparisons are self-contained and do not reduce to the paper's assumptions. The one identified circular element is Section IV-D/Eq. (32), where the R&F 'theoretical analysis' merely restates the κ-scaled learning-rate update as a variance identity and calls it validation; this is a non-central explanatory step. Per the reviewing rule, I also flag explicitly: (a) the Abstract promises 'formulating the resource allocation problem as an Egalitarian Bargaining Game (EBG), our theoretical analysis provides new insights,' but no EBG formulation or theorem appears in Sections III-VI -- an omitted proof, not circularity; (b) Table I omits the reward weights w1-w5, penalty coefficients ωc/ωb/ωbh, stability threshold ε_tol, and coverage threshold Γ_req, and all convergence figures show the R&F trigger fixed at e=500 even though Algorithm 1 defines an adaptive plateau detector -- reproducibility and tuning-artifact risks, but the paper does not explicitly show the trigger was fit to the curves, so I do not count them as demonstrated circularity; (c) self-citations [7], [8], [10] are background deployment results and are not load-bearing. Score 3 rather than 0 because one equation-level reduction is real but limited; not 6 because the central MMF/PF and Pareto-dominance claims do not depend on Eq. (32).

Assumptions & free parameters 7 free parameters · 7 assumptions · 1 invented entities

The central claims rest on a large set of unreported free parameters (reward weights, penalty coefficients, threshold, altitudes) and on several domain assumptions about user distribution, channel model, and interference-free FDMA operation. The one explicit 'theoretical' component, the R&F variance argument, is a definitional consequence of the chosen learning-rate decay rather than an independent result. The EBG promised in the abstract is never instantiated, so the paper contributes a framework plus empirical curves, not the theory it advertises.

free parameters (7)

Reward weights w1–w5 = not reported
Eq. (20) weights coverage, EE, load fairness, rate fairness, and penalty; values are never given but are central to the MMF vs PF comparison.
R&F window size W and stability threshold ε_tol = W=50, ε_tol not reported
Eq. (28)-(29) define plateau detection; ε_tol is absent from Table I and the fixed trigger at episode 500 suggests manual setting.
Learning rate decay factor κ = 0.1
Set in Section IV-D; directly controls the variance-suppression effect claimed in Eq. (32).
Initial UAV altitude h_init = not reported
Used in Eq. (15) for Phase I initialization but not listed in Table I; affects initial path loss and training start.
Coverage threshold Γ_req = not reported
Defines a 'covered' user in Eq. (33); all methods reach ~100% coverage, so the threshold is a discriminating parameter for the reported behavior.
Constraint penalty coefficients ωc, ωb, ωbh = not reported
Eq. (25) penalty terms for collision, boundary, and backhaul violations; values are omitted.
TCP hyperparameters (λp, σ_scatter, mean users per cluster) = only 5 clusters and 50 users are stated
§III-A specifies a Thomas Cluster Process but the intensity and scatter parameters needed to regenerate the exact user distributions are not reported.

assumptions (7)

domain assumption Ground users are distributed as a Thomas Cluster Process
§III-A posits the TCP model to reflect realistic hotspot traffic; all results depend on this synthetic distribution.
domain assumption Air-to-ground LoS probability follows the Al-Hourani model with unspecified urban parameters
Eq. (5)-(8) use S-curve parameters a and b for an 'urban' environment, but the values are not reported.
domain assumption FDMA orthogonal sub-bands remove inter-UAV interference
§III-D assumes each UAV has an exclusive frequency band, so interference is absent from the SNR formulation.
domain assumption Equal-time-sharing (TDMA) within a UAV cell: per-user rate = B/K log2(1+SNR)
Eq. (11) couples user rate to the number of served users, driving the load-balancing behavior.
domain assumption Max-RSSI user association with at most one serving UAV
§III-D defines binary association variables and constraint (13h).
standard math MAPPO clipped surrogate objective and GAE provide a sound policy-gradient update
Eq. (27) is the standard PPO objective; the paper does not prove convergence in this setting.
ad hoc to paper After optimizer reset, the parameter-update variance scales exactly with κ²
Eq. (32) is presented as a mathematical guarantee of R&F, but it assumes the update variance before reset is well-defined and that gradient moments are independent; this is a heuristic claim, not a theorem.

invented entities (1)

Egalitarian Bargaining Game (EBG) formulation
purpose: Claimed in the abstract and intro as the theoretical basis for the fairness-energy-efficiency relationship and for the MMF design rationale.
The EBG appears only in the abstract and introduction; no game-theoretic model, equations, or results are present in the body. It is a named but absent theoretical entity.

how reviews work

0 comments

Cite this review

Pith. "Pith review of ORCHID: Fairness-Aware Orchestration in Mission-Critical Air-Ground Integrated Networks." pith.science (2026). https://pith.science/paper/QYJ3F7X2

@misc{pith2026260209994,
  author       = {Pith},
  title        = {Pith review of: ORCHID: Fairness-Aware Orchestration in Mission-Critical Air-Ground Integrated Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QYJ3F7X2}},
  note         = {Machine review of arXiv:2602.09994}
}

read the original abstract

Unmanned Aerial Vehicles (UAVs) provide pivotal on-demand wireless coverage for mission-critical 6G Air-Ground Integrated Networks (AGINs). However, traditional Deep Reinforcement Learning (DRL) orchestration struggles with multi-agent non-stationarity and balancing Energy Efficiency (EE) with service equity. To address these challenges, we propose ORCHID (Orchestration of Resilient Coverage via Hybrid Intelligent Deployment), a stability-enhanced two-stage learning framework. First, ORCHID utilizes Ground Base Station (GBS)-aware topology partitioning to mitigate the exploration cold-start problem. Second, a Reset-and-Finetune (R&F) mechanism within the Multi-Agent Proximal Policy Optimization (MAPPO) architecture enhances learning stability by synchronizing learning-rate decay with optimizer resetting, thereby reducing gradient variance and mitigating policy degradation. Furthermore, by formulating the resource allocation problem as an Egalitarian Bargaining Game (EBG), our theoretical analysis provides new insights into the relationship between fairness and energy efficiency. Specifically, the proposed Max-Min Fairness (MMF) design provides a theoretical explanation for the emergence of a more dispersed and load-balanced UAV topology, while experimental results further demonstrate that this spatial organization improves system energy efficiency compared with conventional Proportional Fairness (PF) schemes. Moreover, ORCHID deliberately sacrifices opportunistic throughput peaks in favor of more stable long-term service performance, resulting in consistently lower performance variance while maintaining a higher minimum service level and substantially improving service fairness. Extensive experimental results demonstrate robust topology adaptation, stable policy convergence, and consistent performance gains over representative state-of-the-art baselines.

Figures

Figures reproduced from arXiv: 2602.09994 by the authors.

**Figure 2.** The overall architecture of the proposed ORCHID fram [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Training convergence analysis of the ORCHID framewo [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Comparative analysis of convergence performance ag [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]

**Figure 5.** Figure 5: Performance frontier and trade-off analysis. Each d [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

**Figure 6.** Figure 6: Comprehensive ablation study of the ORCHID framewor [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

**Figure 7.** Figure 7: Comparative strategy analysis between ORCHID-MMF a [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 1 linked inside Pith

[1]

Distri- butionally robust optimization for aerial multi-access ed ge computing via cooperation of uavs and haps,

Z. Jia, C. Cui, C. Dong, Q. Wu, Z. Ling, D. Niyato, and Z. Han , “Distri- butionally robust optimization for aerial multi-access ed ge computing via cooperation of uavs and haps,” IEEE Transactions on Mobile Computing, vol. 24, no. 10, pp. 10 853–10 867, 2025

2025
[2]

A survey on drl-based uav communications and networking: D rl fun- damentals, applications and implementations,

W. Zhao, S. Cui, W. Qiu, Z. He, Z. Liu, X. Zheng, B. Mao, and N . Kato, “A survey on drl-based uav communications and networking: D rl fun- damentals, applications and implementations,” IEEE Communications Surveys & Tutorials , vol. 28, pp. 3911–3941, 2026

2026
[3]

A survey on resource management in joint communication and co mputing- embedded sagin,

Q. Chen, Z. Guo, W. Meng, S. Han, C. Li, and T. Q. S. Quek, “A survey on resource management in joint communication and co mputing- embedded sagin,” IEEE Communications Surveys & Tutorials , vol. 27, no. 3, pp. 1911–1954, 2025

1911
[4]

Non-terrestrial networks in t he 6g era: Challenges and opportunities,

M. Giordani and M. Zorzi, “Non-terrestrial networks in t he 6g era: Challenges and opportunities,” IEEE Network , vol. 35, no. 2, pp. 244– 251, 2021

2021
[5]

Joint ua v placement and dependent task ofﬂoading in multi-uav mec net works: a graph attention enhanced drl approach,

C. Zhan, W. Liu, K. Song, R. Fan, J. Liu, and H. Hu, “Joint ua v placement and dependent task ofﬂoading in multi-uav mec net works: a graph attention enhanced drl approach,” IEEE Transactions on Mobile Computing, pp. 1–17, 2025

2025
[6]

Wireless communication s with unmanned aerial vehicles: opportunities and challenges,

Y . Zeng, R. Zhang, and T. J. Lim, “Wireless communication s with unmanned aerial vehicles: opportunities and challenges,” IEEE Com- munications Magazine , vol. 54, no. 5, pp. 36–42, 2016

2016
[7]

The coverage overlapp ing problem of serving arbitrary crowds in 3d drone cellular networks,

C.-C. Lai, L.-C. Wang, and Z. Han, “The coverage overlapp ing problem of serving arbitrary crowds in 3d drone cellular networks,” IEEE Transactions on Mobile Computing , vol. 21, no. 3, 2022

2022
[8]

On-demand densit y-aware uav base station 3d placement for arbitrarily distributed u sers with guaranteed data rates,

C.-C. Lai, C.-T. Chen, and L.-C. Wang, “On-demand densit y-aware uav base station 3d placement for arbitrarily distributed u sers with guaranteed data rates,” IEEE Wireless Communications Letters , vol. 8, no. 3, pp. 913–916, 2019

2019

Show all 34 references

[9]

J oint optimization of 3d placement and radio resource allocation for per-uav sum rate maximization,

A. Mahmood, T. X. Vu, S. Chatzinotas, and B. Ottersten, “J oint optimization of 3d placement and radio resource allocation for per-uav sum rate maximization,” IEEE Transactions on V ehicular Technology , vol. 72, no. 10, pp. 13 094–13 105, 2023

2023
[10]

Adaptive and fair deployment approach to balance ofﬂoad trafﬁc in multi-uav c ellular networks,

C.-C. Lai, Bhola, A.-H. Tsai, and L.-C. Wang, “Adaptive and fair deployment approach to balance ofﬂoad trafﬁc in multi-uav c ellular networks,” IEEE Transactions on V ehicular Technology, vol. 72, no. 3, pp. 3724–3738, 2023

2023
[11]

Aquantitative measur e of fairness and discrimination for resource allocation in shared syste ms,

R. Jain, D.-M. Chiu, and W. Hawe, “Aquantitative measur e of fairness and discrimination for resource allocation in shared syste ms,” digital equipment corporation,” Technical Report DEC-TR-301, Tec h. Rep., 1984

1984
[12]

Uav swarm cooperative dynamic target search: A mappo-based discrete optimal control method,

D. Wei, L. Zhang, Q. Liu, H. Chen, and J. Huang, “Uav swarm cooperative dynamic target search: A mappo-based discrete optimal control method,” Drones, vol. 8, no. 6, 2024

2024
[13]

Multi- agent actor-critic for mixed cooperative-competitive env ironments,

R. Lowe, Y . Wu, A. Tamar, J. Harb, P . Abbeel, and I. Mordat ch, “Multi- agent actor-critic for mixed cooperative-competitive env ironments,” in Proceedings of the 31st International Conference on Neural Information Processing Systems , Long Beach, California, USA, 2017, pp. ...

2017
[14]

The surprising effectiveness of ppo in cooperative multi-agen t games,

C. Y u, A. V elu, E. Vinitsky, J. Gao, Y . Wang, A. Bayen, and Y . Wu, “The surprising effectiveness of ppo in cooperative multi-agen t games,” in Proceedings of the 36th International Conference on Neural Information Processing Systems, New Orleans, LA, USA, 2022

2022
[15]

k-means++: the advant ages of careful seeding,

D. Arthur and S. V assilvitskii, “k-means++: the advant ages of careful seeding,” in Proceedings of the Eighteenth Annual ACM-SIAM Sympo- sium on Discrete Algorithms , New Orleans, Louisiana, 2007, pp. 1027– 1035

2007
[16]

Efﬁcie nt de- ployment of multiple unmanned aerial vehicles for optimal w ireless coverage,

M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, “Efﬁcie nt de- ployment of multiple unmanned aerial vehicles for optimal w ireless coverage,” IEEE Communications Letters, vol. 20, no. 8, pp. 1647–1650, 2016

2016
[17]

Optimizing uav deployment for maximi zing coverage and data rate efﬁciency using multi-agent deep det erministic policy gradient and bayesian optimization,

D. K. R. and R. A., “Optimizing uav deployment for maximi zing coverage and data rate efﬁciency using multi-agent deep det erministic policy gradient and bayesian optimization,” Physical Communication , vol. 69, p. 102621, 2025

2025
[18]

An optimized 2d ground area coverage using uav-enabled sen sor networks,

S. Swain, R. Ranjan Swain, B. Ranjan Senapati, and P . Moh an Khilar, “An optimized 2d ground area coverage using uav-enabled sen sor networks,” IEEE Access , vol. 13, pp. 161 299–161 310, 2025

2025
[19]

Edge-prior placemen t algorithm for uav-mounted base stations,

J. Qin, Z. Wei, C. Qiu, and Z. Feng, “Edge-prior placemen t algorithm for uav-mounted base stations,” in 2019 IEEE Wireless Communications and Networking Conference (WCNC) , 2019, pp. 1–6

2019
[20]

Securing uav commu nications via joint trajectory and power control,

G. Zhang, Q. Wu, M. Cui, and R. Zhang, “Securing uav commu nications via joint trajectory and power control,” IEEE Transactions on Wireless Communications, vol. 18, no. 2, pp. 1376–1389, 2019

2019
[21]

Noma for star-ris ass isted uav networks,

J. Lei, T. Zhang, X. Mu, and Y . Liu, “Noma for star-ris ass isted uav networks,” IEEE Transactions on Communications , vol. 72, no. 3, pp. 1732–1745, 2024

2024
[22]

Uav-aided multi-way noma networks with residual hardware impair- ments,

X. Li, Q. Wang, Y . Liu, T. A. Tsiftsis, Z. Ding, and A. Nall anathan, “Uav-aided multi-way noma networks with residual hardware impair- ments,” IEEE Wireless Communications Letters , vol. 9, no. 9, pp. 1538– 1542, 2020

2020
[23]

Distributed energy -efﬁcient multi-uav navigation for long-term communication coverag e by deep reinforcement learning,

C. H. Liu, X. Ma, X. Gao, and J. Tang, “Distributed energy -efﬁcient multi-uav navigation for long-term communication coverag e by deep reinforcement learning,” IEEE Transactions on Mobile Computing , vol. 19, no. 6, pp. 1274–1285, 2020

2020
[24]

Coop erative internet of uavs: Distributed trajectory design by multi-a gent deep re- inforcement learning,

J. Hu, H. Zhang, L. Song, R. Schober, and H. V . Poor, “Coop erative internet of uavs: Distributed trajectory design by multi-a gent deep re- inforcement learning,” IEEE Transactions on Communications , vol. 68, no. 11, pp. 6807–6821, 2020

2020
[25]

Multi- agent deep reinforcement learning based uav trajectory opt imization for differentiated services,

Z. Ning, Y . Y ang, X. Wang, Q. Song, L. Guo, and A. Jamalipo ur, “Multi- agent deep reinforcement learning based uav trajectory opt imization for differentiated services,” IEEE Transactions on Mobile Computing , vol. 23, no. 5, pp. 5818–5834, 2024

2024
[26]

Cooperative uav resource allocation and task ofﬂoading in hierarchical aerial computing systems: A mappo-based approach,

H. Kang, X. Chang, J. Miˇ si´ c, V . B. Miˇ si´ c, J. Fan, and Y . Liu, “Cooperative uav resource allocation and task ofﬂoading in hierarchical aerial computing systems: A mappo-based approach,” IEEE Internet of Things Journal, vol. 10, no. 12, pp. 10 497–10 509, 2023

2023
[27]

Joint power allo cation and 3d deployment for uav-bss: A game theory based deep reinforc ement learning approach,

S. Fu, X. Feng, A. Sultana, and L. Zhao, “Joint power allo cation and 3d deployment for uav-bss: A game theory based deep reinforc ement learning approach,” IEEE Transactions on Wireless Communications , vol. 23, no. 1, pp. 736–748, 2024

2024
[28]

Fairness -aware task scheduling and resource allocation in uav-enabled mob ile edge computing networks,

M. Zhao, W. Li, L. Bao, J. Luo, Z. He, and D. Liu, “Fairness -aware task scheduling and resource allocation in uav-enabled mob ile edge computing networks,” IEEE Transactions on Green Communications and Networking , vol. 5, no. 4, pp. 2174–2187, 2021

2021
[29]

S. N. Chiu, D. Stoyan, W. S. Kendall, and J. Mecke, Stochastic Geometry and its Applications . John Wiley & Sons, 2013

2013
[30]

Optimal l ap altitude for maximum coverage,

A. Al-Hourani, S. Kandeepan, and S. Lardner, “Optimal l ap altitude for maximum coverage,” IEEE Wireless Communications Letters , vol. 3, no. 6, pp. 569–572, 2014

2014
[31]

Dealing with non-stationarity in multi-agent deep reinforcement l earning,

G. Papoudakis, F. Christianos, A. Rahman, and S. V . Albr echt, “Dealing with non-stationarity in multi-agent deep reinforcement l earning,” 2019. [Online]. Available: https://arxiv.org/abs/1906.04737 SUBMITTED TO IEEE JOURNAL FOR POSSIBLE PUBLICA TION, JANUAR Y 2026 15

2019 arXiv
[32]

The primacy bias in deep reinforcement learning,

E. Nikishin, M. Schwarzer, P . D’Oro, P .-L. Bacon, and A. Courville, “The primacy bias in deep reinforcement learning,” in International Conference on Machine Learning (ICML) , Baltimore, Maryland, USA, 2022

2022
[33]

On warm-starting neural netwo rk training,

J. T. Ash and R. P . Adams, “On warm-starting neural netwo rk training,” in International Conference on Neural Information Processin g Systems (NIPS), V ancouver, BC, Canada, 2020

2020
[34]

On the convergence of adam and beyond,

S. J. Reddi, S. Kale, and S. Kumar, “On the convergence of adam and beyond,” in International Conference on Learning Representations (ICLR), V ancouver, BC, Canada, 2018. Chuan-Chi Lai (Member, IEEE) received the Ph.D. degree in Computer Science and Information Engi- neering f...

2018

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Distri- butionally robust optimization for aerial multi-access ed ge computing via cooperation of uavs and haps,

Z. Jia, C. Cui, C. Dong, Q. Wu, Z. Ling, D. Niyato, and Z. Han , “Distri- butionally robust optimization for aerial multi-access ed ge computing via cooperation of uavs and haps,” IEEE Transactions on Mobile Computing, vol. 24, no. 10, pp. 10 853–10 867, 2025

2025

[2] [2]

A survey on drl-based uav communications and networking: D rl fun- damentals, applications and implementations,

W. Zhao, S. Cui, W. Qiu, Z. He, Z. Liu, X. Zheng, B. Mao, and N . Kato, “A survey on drl-based uav communications and networking: D rl fun- damentals, applications and implementations,” IEEE Communications Surveys & Tutorials , vol. 28, pp. 3911–3941, 2026

2026

[3] [3]

A survey on resource management in joint communication and co mputing- embedded sagin,

Q. Chen, Z. Guo, W. Meng, S. Han, C. Li, and T. Q. S. Quek, “A survey on resource management in joint communication and co mputing- embedded sagin,” IEEE Communications Surveys & Tutorials , vol. 27, no. 3, pp. 1911–1954, 2025

1911

[4] [4]

Non-terrestrial networks in t he 6g era: Challenges and opportunities,

M. Giordani and M. Zorzi, “Non-terrestrial networks in t he 6g era: Challenges and opportunities,” IEEE Network , vol. 35, no. 2, pp. 244– 251, 2021

2021

[5] [5]

Joint ua v placement and dependent task ofﬂoading in multi-uav mec net works: a graph attention enhanced drl approach,

C. Zhan, W. Liu, K. Song, R. Fan, J. Liu, and H. Hu, “Joint ua v placement and dependent task ofﬂoading in multi-uav mec net works: a graph attention enhanced drl approach,” IEEE Transactions on Mobile Computing, pp. 1–17, 2025

2025

[6] [6]

Wireless communication s with unmanned aerial vehicles: opportunities and challenges,

Y . Zeng, R. Zhang, and T. J. Lim, “Wireless communication s with unmanned aerial vehicles: opportunities and challenges,” IEEE Com- munications Magazine , vol. 54, no. 5, pp. 36–42, 2016

2016

[7] [7]

The coverage overlapp ing problem of serving arbitrary crowds in 3d drone cellular networks,

C.-C. Lai, L.-C. Wang, and Z. Han, “The coverage overlapp ing problem of serving arbitrary crowds in 3d drone cellular networks,” IEEE Transactions on Mobile Computing , vol. 21, no. 3, 2022

2022

[8] [8]

On-demand densit y-aware uav base station 3d placement for arbitrarily distributed u sers with guaranteed data rates,

C.-C. Lai, C.-T. Chen, and L.-C. Wang, “On-demand densit y-aware uav base station 3d placement for arbitrarily distributed u sers with guaranteed data rates,” IEEE Wireless Communications Letters , vol. 8, no. 3, pp. 913–916, 2019

2019