GFlowNets for Active Learning Based Resource Allocation in Next Generation Wireless Networks

Charbel Bou Chaaya; Mehdi Bennis

REVIEW 5 major objections 4 minor 23 references

GFlowNets for Active Learning Based Resource Allocation in Next Generation Wireless Networks

T0 review · 5 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read An active learning loop pairing a GFlowNet sampler with a Gaussian-process surrogate allocates radio resources with 20% higher utility in less than half the acquisition rounds of PPO or MCMC.

desk verdict A plausible new application of GFlowNet active learning to ISCC resource allocation, but the headline gains are not verifiable as written because the generator's stated constraints do not enforce the system model's one-block-per-device rule. read the letter →

arxiv 2505.05224 v1 pith:4E5NMD3O submitted 2025-05-08 cs.LG

Charbel Bou Chaaya , Mehdi Bennis This is my paper · ORCID

classification cs.LG

keywords RadioresourcemanagementIntegratedsensingcommunicationandcomputingActivelearningGenerativeflownetworkGaussianprocesssurrogateTrajectorybalanceallocationWirelessnetworks

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 targets radio resource allocation in wireless networks that must simultaneously serve communication, sensing, and computing traffic. The authors formulate the problem as maximizing a sum of per-device utilities and observe that the resulting search over allocation matrices is NP-hard and high-dimensional. They propose a generative active learning loop: a Gaussian process models the expensive true utility from past allocations, and a GFlowNet—trained so that sampled matrices are proportional to the surrogate reward—proposes the next set of allocations to evaluate. The paper reports that this loop reaches higher-utility allocations in about half the acquisition rounds of PPO or MCMC and keeps a roughly 20% advantage as the number of devices grows. A sympathetic reader would care because it suggests a practical, general recipe for quickly configuring multi-functional wireless networks at the start of each scheduling slot.

What carries the argument

The load-bearing mechanism is the combination of a Gaussian-process surrogate with a GFlowNet generator inside a sequential active learning loop. A GFlowNet is a generative model over compositional objects: it builds an M×W allocation matrix one entry at a time from an all-zero start, with each action setting X_{m,w}=d, and it is trained with the trajectory balance loss so that terminal objects are sampled with probability proportional to the reward—here the GP's predicted utility. The GP uses a Matérn kernel applied to neural-network embeddings of the matrices, keeping the surrogate tractable. The paper argues that proportional-to-reward sampling yields better mode coverage than PPO's expected-return maximization and faster mixing than MCMC, which is what allows fewer acquisition rounds.

What would settle it

Sample a few thousand terminal matrices from the trained GFlowNet and check the system model's per-device constraint: if the same device appears in multiple resource blocks while some other device appears in none, the claimed feasibility guarantee fails and the reported utilities are computed over invalid allocations. A simpler version: re-run the active loop with an explicit one-block-per-device constraint; if the 20% gain disappears, the original advantage depended on those invalid configurations.

Watch

Extended reading notes

Core claim

The paper's central claim is that a generative flow network trained with the trajectory balance objective can serve as the sampler in an active learning loop for combinatorial radio resource allocation, outperforming both a deep RL policy (PPO) and a classical MCMC sampler. The utility optimized is the sum over devices of weighted products of communication bitrate, sensing mutual information, and inverse computing latency, which couples MEC-server selection and subcarrier assignment. In simulations with 20 devices, the GFlowNet agent reaches its best solution in roughly five acquisition rounds—about half the ten rounds PPO needs—with a solution about 8% higher, while MCMC takes about twenty rounds for a solution 4% above PPO. Scaled over device counts up to 40, the GFlowNet consistently outperforms both baselines by about 20% after ten rounds. The authors directly attribute the gain to GFlowNet's training objective, which makes the probability of sampling an allocation proportional to its surrogate reward and therefore keeps exploring diverse high-return modes.

Load-bearing premise

The central claim assumes the GFlowNet's stated constraints—never reusing a resource block and editing exactly as many entries as there are devices—guarantee feasible allocations, but those rules still allow one device to take several blocks while another device gets none, and the paper does not say whether device-uniqueness is enforced.

Editorial extensions

If this is right

If the reported gains are correct, a GFlowNet-based scheduler can configure a multi-functional wireless network at the start of each slot with less than half the system evaluations required by PPO or MCMC.
Because the trajectory balance objective makes sampling proportional to the surrogate reward, the active loop keeps proposing diverse high-utility allocations rather than collapsing onto one optimum, which is the failure mode the paper attributes to PPO.
Applied to other combinatorial radio resource problems with expensive objectives—beamforming pattern selection, user association, subcarrier assignment—the same GFlowNet-plus-GP loop would need only a redefinition of the DAG's actions.
The reduction in acquisition rounds translates directly into shorter scheduling delay in the time-slotted scenario the paper assumes, since the allocation is fixed for the duration of the slot.

Reading between the lines

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

A clean control that keeps the GP surrogate but replaces the GFlowNet with random sampling would isolate whether the reported gains come from the sampler's proportional-to-reward coverage or simply from the GP learning; the paper does not report such an ablation.
Since the generator's stated constraints do not explicitly stop one device from taking several resource blocks while a neighbor gets none, the natural robustness check is to add a one-block-per-device constraint and re-measure the 20% gain; the paper leaves that check unstated.
The conclusion names multi-fidelity active learning as future work; a concrete version would let the sampler choose the evaluation fidelity, spending cheap surrogate evaluations early and expensive oracle evaluations only around promising modes.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Quick take: this is a legitimate new application—GFlowNet-based active learning for joint communication, sensing, and computing resource allocation—and the problem formulation is sensible. But the paper as written does not support its headline claims. The generator model in Section IV says that forbidding reuse of a block (m,w) and setting the number of modified entries to U "guarantees" feasible allocations. That guarantee is false. Nothing stops the same device from occupying two blocks while another device gets none, which violates Section III's rule that each device is allocated to one MEC server over one block. That is not a cosmetic detail: if the implementation does not mask already-assigned devices, the reported utilities can be computed over infeasible matrices, artificially inflating throughput and sensing for duplicated devices and zeroing the utility for omitted ones. If the implementation does enforce device uniqueness, the text is simply incomplete—but either way, the abstract's 20% gain and half-round claim are not verifiable from the manuscript.

What the paper does well: the active learning loop (GP surrogate plus GFlowNet sampler) is a coherent way to handle an expensive, discrete, high-dimensional utility, and the comparison against PPO and MCMC is a reasonable sanity check. The multi-objective utility product in (12) is a fair way to capture communication, sensing, and computing trade-offs. The simulation setup is clearly described, and the figures suggest the method finds higher-utility solutions faster than the baselines, at least in the scenarios shown.

Soft spots, in proportion: the feasibility gap is load-bearing and needs to be fixed. There are no error bars, no seeds, and no code, so the empirical results are hard to trust even after the feasibility issue is resolved. The abstract combines two separate results—Fig. 3a's 8% gain in convergence speed and Fig. 3b's ~20% gain in scalability—into one "20% gains while requiring less than half the rounds" headline, which is misleading. The sensing SINR expression in (10) also looks inconsistent: the numerator sums over MEC m while the denominator excludes m' = m, so the interference model is not clearly specified. Minor: the novelty is mostly the application; the GFlowNet and active learning machinery is off-the-shelf, and the authors should say that more plainly.

Who this is for: researchers working on data-driven radio resource management, especially those interested in active learning or generative samplers for combinatorial allocation. A serious referee should engage with this, but the paper needs major revision before the empirical claims can be accepted. If the authors add the missing device-uniqueness constraint, report multiple seeds with standard deviations, and separate the convergence and scalability claims, the core idea could be a useful contribution.

Recommendation: send to peer review, but with the expectation of substantial revision.

Referee Report

5 major / 4 minor

Summary. The paper proposes an active-learning framework for radio resource allocation in an integrated sensing, communication, and computing network. A Gaussian process surrogate is iteratively updated with utility evaluations, and a GFlowNet is trained against this surrogate to propose diverse high-return resource allocation matrices. The authors report that the method achieves about 20% utility gains over PPO and MCMC baselines in a scalability experiment, and reaches the baseline performance in about half the acquisition rounds in a convergence experiment. The central novelty is the use of a GFlowNet as a discrete compositional sampler inside an active-learning loop.

Significance. If the reported results hold, the paper makes a useful contribution by showing that GFlowNet-based generative active learning can be a viable heuristic for high-dimensional discrete resource allocation, where classical RL and MCMC suffer from mode collapse or slow mixing. The evaluation is not circular: the GFlowNet is trained on the GP surrogate while the reported utilities are computed from the true system model, so the numbers are not an artifact of the surrogate. The main strengths are the clear problem formulation, the use of trajectory balance with a stated convergence guarantee, and a sensible comparison setup. However, the current manuscript leaves the central empirical claims unverifiable because of an incorrect feasibility argument, an ill-defined sensing SINR expression, and missing experimental details such as seeds, error bars, and hyperparameter values.

major comments (5)

[Section IV, Generator Model] The feasibility guarantee stated in Section IV does not follow from the constraints described. The text says that previously selected resource blocks cannot be re-used and that the number of modified entries equals U, then concludes that terminal objects are feasible allocation matrices. Section III defines feasibility as each device being allocated to one MEC server over one resource block. The two stated constraints only ensure that U distinct entries are non-zero; they do not prevent the same device from occupying several blocks while another device receives none. Such a matrix violates the system model, and the computing latency in (11) is undefined for an omitted device because the numerator has no allocated block. The fix is to add an explicit constraint in the action mask forbidding a device that has already been assigned, and to state that constraint in the text. If the implementation silently enforces this, the description is incomplete; if it does not, the reported utilities may be computed over infeasible allocations.
[Abstract and Section V, Figs. 3a and 3b] The abstract combines two separate gains into one claim. Figure 3a shows that GFlowNet reaches a solution about 8% better than PPO in half the acquisition rounds (5 versus 10), while Figure 3b shows roughly 20% improvement after 10 rounds when the number of devices is varied. As written, the abstract's statement '20% performance gains against benchmarks, while requiring less than half of the number of acquisition rounds' implies the 20% gain is achieved at the half-round point, which is not supported by the displayed curves. Please report these two results separately, or show a joint trade-off curve so that the claim matches the evidence.
[Section V, Eq. (10)] The sensing SINR in Eq. (10) is not well defined. The numerator sums over the MEC index m, but the denominator contains the term m' ≠ m with m appearing as a free variable outside the numerator's summation convention. As typeset, the expression has no unambiguous meaning. Moreover, if the generator can produce a matrix in which device d occupies multiple blocks, the numerator would sum over multiple terms, which is inconsistent with the one-block-per-device rule of Section III. Please rewrite Eq. (10) with a fixed MEC index or a clear convention, and ensure it is consistent with the feasibility constraints.
[Section V, Simulation Results] No random seeds, number of independent runs, or error bars are reported for any figure. GFlowNet, PPO, and MCMC are all stochastic algorithms, and the text makes strong statements such as 'consistently outperforming benchmarks by around 20%' based on what appears to be a single run. Please add the number of seeds, report mean and standard deviation (or confidence intervals), and describe the variance across runs. This is necessary to verify the robustness of the central claim.
[Section V, Simulation Setup] Several parameters that are essential for reproducibility are missing: the utility weights ω_d^ρ, ω_d^ζ, and ω_d^ξ in Eq. (12) are not specified; the GP kernel hyperparameters ν and l, and the architecture and training details of the deep kernel network, are not reported; and the number of GFlowNet training steps per active-learning round and the backward policy parameterization are not given. Without these details, the reported gains could depend on unstated tuning choices. Please provide the full experimental configuration.

minor comments (4)

[Section II-A] In the definition of terminal states, the phrase 'states with no outgoing edges are terminal states are terminal states' contains a duplicated fragment. Please revise.
[Section III] The feasibility rule 'each device can be allocated to communicate with one MEC server over one resource block; conversely, each subcarrier can be retained for one device only' is stated informally. It would be clearer to define the set X of feasible allocations explicitly, including whether a device must receive exactly one block or at most one block.
[Section III-D] The claim that problem (13) is 'a variant of the knapsack problem which belongs to the class of NP-complete problems' is stated without a reduction or reference. A brief justification or citation would be appropriate.
[Section V, Fig. 4] The claim that the GFlowNet solution is 'much better' than the baselines is not quantified. Please add numerical values or a table, since the 3D scatter plot alone does not show the utility differences clearly.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: GFlowNet proposals are evaluated on the true utility, not on the surrogate they were trained against.

full rationale

The paper's central claim is an optimization result: a GFlowNet trained against a GP surrogate in an active-learning loop proposes resource allocations whose utilities are then evaluated by the true environment oracle f in (13). The reported gains are therefore not measured against the same function used to train or reward the sampler. The GP surrogate h is fitted to evaluations of f, and the GFlowNet is trained to sample proportionally to h, but the final utilities in Figs. 3 and 4 are computed from the system model f via (5)-(12), not from h, so no fitted parameter is renamed as a prediction. The GFlowNet references [14], [17], and [18] are external foundations used as tools; the only author self-citation ([2], supporting the broad statement that classical resource optimization involves zero-sum trade-offs) is motivational and does not carry the derivation. The feasibility gap noted in Section IV (the stated constraints do not by themselves prevent a device from being assigned to several blocks while another receives none) is a correctness or implementation-verifiability risk, but it is not circularity: it does not make the reported utility equal to the method's input by construction. No circular step meets the evidentiary bar of the review.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The method depends on standard GFlowNet convergence and GP regression results, on several domain modeling choices for sensing and computing, and on hand-chosen objective weights that are never reported. The only formally fragile structural element is the feasibility guarantee for the generative model, which is addressed under red flags.

free parameters (3)

Utility weights omega_d^rho, omega_d^zeta, omega_d^xi = unspecified
These positive weights define the scalar objective in Eq. (12) and therefore determine which allocations count as high utility. The paper never states their values, so the reported utility numbers cannot be reproduced or compared across scenarios.
GP kernel hyperparameters nu and l (deep kernel) = optimized internally
Used in the Matérn kernel in Eq. (14) and learned from data via deep kernel learning; values are not reported, which affects the surrogate quality and is part of the method's tuning burden.
GFlowNet training steps per round and backward policy details = not stated
The policy MLP architecture (two hidden layers of 64 ReLU units) and learning rate 0.001 are given, but the number of training iterations per acquisition round, trajectory batch size, and backward policy parametrization are not specified; these hand-chosen implementation choices affect performance.

assumptions (6)

standard math Trajectory balance objective converges so that sampled objects are proportional to reward (Proposition 1 of [18]).
The paper relies on this cited theorem to justify the GFlowNet sampling property; it is not re-derived here.
standard math GP posterior inference equations (3)-(4) provide a valid surrogate of the unknown utility.
Standard Gaussian process regression result used for the surrogate model.
domain assumption The resource allocation problem (13) is NP-complete as a variant of the knapsack problem.
Asserted without proof; even if false, the active-learning approach is still motivated by computational expense.
domain assumption Channels are static during each scheduling slot and independent across slots.
Justifies solving the allocation at the start of each slot; entered in Section III.D.
domain assumption Sensing performance is characterized by conditional mutual information between target impulse response and received signal.
Model borrowed from [21]; underlies the sensing utility in Eqs. (9)-(10).
domain assumption MEC server capacity is uniformly split among served devices.
Computing model in Eq. (11); used to define the latency metric.

how reviews work

0 comments

Cite this review

Pith. "Pith review of GFlowNets for Active Learning Based Resource Allocation in Next Generation Wireless Networks." pith.science (2026). https://pith.science/paper/4E5NMD3O

@misc{pith2026250505224,
  author       = {Pith},
  title        = {Pith review of: GFlowNets for Active Learning Based Resource Allocation in Next Generation Wireless Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4E5NMD3O}},
  note         = {Machine review of arXiv:2505.05224}
}

read the original abstract

In this work, we consider the radio resource allocation problem in a wireless system with various integrated functionalities, such as communication, sensing and computing. We design suitable resource management techniques that can simultaneously cater to those heterogeneous requirements, and scale appropriately with the high-dimensional and discrete nature of the problem. We propose a novel active learning framework where resource allocation patterns are drawn sequentially, evaluated in the environment, and then used to iteratively update a surrogate model of the environment. Our method leverages a generative flow network (GFlowNet) to sample favorable solutions, as such models are trained to generate compositional objects proportionally to their training reward, hence providing an appropriate coverage of its modes. As such, GFlowNet generates diverse and high return resource management designs that update the surrogate model and swiftly discover suitable solutions. We provide simulation results showing that our method can allocate radio resources achieving 20% performance gains against benchmarks, while requiring less than half of the number of acquisition rounds.

Figures

Figures reproduced from arXiv: 2505.05224 by the authors.

**Figure 1.** System model. active learning rounds proceed as follows: the agent initially observes a dataset D0 = x 0 i , y0 i n i=1, where y 0 i = f [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 2.** GFlowNet model to sample resource allocation schemes (commu [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Comparison between different algorithms. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: A sample of solutions found by different methods. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 19 canonical work pages

[1]

Twelve scientific challenges for 6g: Rethink- ing the foundations of communications theory,

M. Chafii, L. Bariah et al., “Twelve scientific challenges for 6g: Rethink- ing the foundations of communications theory,” IEEE Communications Surveys & Tutorials, vol. 25, no. 2, pp. 868–904, 2023

work page 2023
[2]

Ultrareliable and low-latency wireless communication: Tail, risk, and scale,

M. Bennis, M. Debbah, and H. V . Poor, “Ultrareliable and low-latency wireless communication: Tail, risk, and scale,” Proceedings of the IEEE, vol. 106, no. 10, pp. 1834–1853, 2018

2018
[3]

A survey on integrated sensing, communication, and computation,

D. Wen, Y . Zhou et al., “A survey on integrated sensing, communication, and computation,” IEEE Communications Surveys & Tutorials , 2024

work page 2024
[4]

Radio resource allocation for integrated sensing, communication, and computation networks,

L. Zhao, D. Wu et al., “Radio resource allocation for integrated sensing, communication, and computation networks,” IEEE Transactions on Wireless Communications, vol. 21, no. 10, pp. 8675–8687, 2022

work page 2022
[5]

Integrated sensing, computation, and com- munication: System framework and performance optimization,

Y . He, G. Yu et al. , “Integrated sensing, computation, and com- munication: System framework and performance optimization,” IEEE Transactions on Wireless Communications , vol. 23, no. 2, pp. 1114– 1128, 2023

work page 2023
[6]

Integrated sensing, communication, and compu- tation over-the-air: Mimo beamforming design,

X. Li, F. Liu et al. , “Integrated sensing, communication, and compu- tation over-the-air: Mimo beamforming design,” IEEE Transactions on Wireless Communications, vol. 22, no. 8, pp. 5383–5398, 2023

work page 2023
[7]

Integrated sensing-communication- computation for over-the-air edge ai inference,

Z. Zhuang, D. Wen et al. , “Integrated sensing-communication- computation for over-the-air edge ai inference,” IEEE Transactions on Wireless Communications, vol. 23, no. 4, pp. 3205–3220, 2023

work page 2023
[8]

Mean field game-based waveform precoding design for mobile crowd integrated sensing, communication, and compu- tation systems,

D. Wang, C. Huang et al., “Mean field game-based waveform precoding design for mobile crowd integrated sensing, communication, and compu- tation systems,” IEEE Transactions on Wireless Communications, 2024

work page 2024

Show all 23 references

[9]

Joint sensing, communication, and computation resource allocation for cooperative perception in fog-based vehicular networks,

X. Zhang, Z. He et al., “Joint sensing, communication, and computation resource allocation for cooperative perception in fog-based vehicular networks,” in 2021 13th International Conference on Wireless Commu- nications and Signal Processing (WCSP) . IEEE, 2021, pp. 1–6

2021
[10]

Deep reinforcement learning-based resource allocation for integrated sensing, communication, and computation in vehicular network,

L. Yang, Y . Wei et al. , “Deep reinforcement learning-based resource allocation for integrated sensing, communication, and computation in vehicular network,” IEEE Transactions on Wireless Communications , 2024

2024
[11]

Gdsg: Graph diffusion-based solution gen- eration for optimization problems in mec networks,

R. Liang, B. Yang et al. , “Gdsg: Graph diffusion-based solution gen- eration for optimization problems in mec networks,” arXiv preprint arXiv:2412.08296, 2024

2024 arXiv
[12]

Active learning: A survey,

C. C. Aggarwal, X. Kong et al. , “Active learning: A survey,” in Data classification. Chapman and Hall/CRC, 2014, pp. 599–634

2014
[13]

Bayesian optimization for radio resource management: Open loop power control,

L. Maggi, A. Valcarce, and J. Hoydis, “Bayesian optimization for radio resource management: Open loop power control,” IEEE Journal on Selected Areas in Communications, vol. 39, no. 7, pp. 1858–1871, 2021

2021
[14]

Flow network based generative models for non-iterative diverse candidate generation,

E. Bengio, M. Jain et al. , “Flow network based generative models for non-iterative diverse candidate generation,” Advances in Neural Information Processing Systems , vol. 34, pp. 27 381–27 394, 2021

2021
[15]

Biological sequence design with gflownets,

M. Jain, E. Bengio et al., “Biological sequence design with gflownets,” in International Conference on Machine Learning . PMLR, 2022, pp. 9786–9801

2022
[16]

Multi-fidelity active learning with gflownets,

A. Hernández-García, N. Saxena et al. , “Multi-fidelity active learning with gflownets,” Transactions on Machine Learning Research , 2024

2024
[17]

Gflownet foundations,

Y . Bengio, S. Lahlou et al., “Gflownet foundations,” Journal of Machine Learning Research, vol. 24, no. 210, pp. 1–55, 2023

2023
[18]

Trajectory balance: Improved credit assign- ment in gflownets,

N. Malkin, M. Jain et al., “Trajectory balance: Improved credit assign- ment in gflownets,” Advances in Neural Information Processing Systems, vol. 35, pp. 5955–5967, 2022

2022
[19]

From theories to queries: Active learning in practice,

B. Settles, “From theories to queries: Active learning in practice,” in Active learning and experimental design workshop in conjunction with AISTATS 2010. JMLR Workshop and Conference Proceedings, 2011, pp. 1–18

2010
[20]

C. E. Rasmussen and C. K. I. Williams, Gaussian Processes for Machine Learning. The MIT Press, 11 2005

2005
[21]

Adaptive ofdm integrated radar and com- munications waveform design based on information theory,

Y . Liu, G. Liao et al. , “Adaptive ofdm integrated radar and com- munications waveform design based on information theory,” IEEE Communications Letters, vol. 21, no. 10, pp. 2174–2177, 2017

2017
[22]

Stochastic variational deep kernel learning,

A. G. Wilson, Z. Hu et al., “Stochastic variational deep kernel learning,” Advances in neural information processing systems , vol. 29, 2016

2016
[23]

Proximal policy optimization algorithms,

J. Schulman, F. Wolski et al., “Proximal policy optimization algorithms,” arXiv preprint arXiv:1707.06347 , 2017

2017 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Twelve scientific challenges for 6g: Rethink- ing the foundations of communications theory,

M. Chafii, L. Bariah et al., “Twelve scientific challenges for 6g: Rethink- ing the foundations of communications theory,” IEEE Communications Surveys & Tutorials, vol. 25, no. 2, pp. 868–904, 2023

work page 2023

[2] [2]

Ultrareliable and low-latency wireless communication: Tail, risk, and scale,

M. Bennis, M. Debbah, and H. V . Poor, “Ultrareliable and low-latency wireless communication: Tail, risk, and scale,” Proceedings of the IEEE, vol. 106, no. 10, pp. 1834–1853, 2018

2018

[3] [3]

A survey on integrated sensing, communication, and computation,

D. Wen, Y . Zhou et al., “A survey on integrated sensing, communication, and computation,” IEEE Communications Surveys & Tutorials , 2024

work page 2024

[4] [4]

Radio resource allocation for integrated sensing, communication, and computation networks,

L. Zhao, D. Wu et al., “Radio resource allocation for integrated sensing, communication, and computation networks,” IEEE Transactions on Wireless Communications, vol. 21, no. 10, pp. 8675–8687, 2022

work page 2022

[5] [5]

Integrated sensing, computation, and com- munication: System framework and performance optimization,

Y . He, G. Yu et al. , “Integrated sensing, computation, and com- munication: System framework and performance optimization,” IEEE Transactions on Wireless Communications , vol. 23, no. 2, pp. 1114– 1128, 2023

work page 2023

[6] [6]

Integrated sensing, communication, and compu- tation over-the-air: Mimo beamforming design,

X. Li, F. Liu et al. , “Integrated sensing, communication, and compu- tation over-the-air: Mimo beamforming design,” IEEE Transactions on Wireless Communications, vol. 22, no. 8, pp. 5383–5398, 2023

work page 2023

[7] [7]

Integrated sensing-communication- computation for over-the-air edge ai inference,

Z. Zhuang, D. Wen et al. , “Integrated sensing-communication- computation for over-the-air edge ai inference,” IEEE Transactions on Wireless Communications, vol. 23, no. 4, pp. 3205–3220, 2023

work page 2023

[8] [8]

Mean field game-based waveform precoding design for mobile crowd integrated sensing, communication, and compu- tation systems,

D. Wang, C. Huang et al., “Mean field game-based waveform precoding design for mobile crowd integrated sensing, communication, and compu- tation systems,” IEEE Transactions on Wireless Communications, 2024

work page 2024