Pith. sign in

REVIEW 4 major objections 4 minor 44 references

Heterogeneous Resource Allocation with Multi-task Learning for Wireless Networks

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A single deep network, guided by a small router that learns a binary mask per task, can jointly solve wireless resource-allocation problems that differ in size, objective, and constraints, nearly matching separately trained single-task…

desk verdict A useful application paper that extends prior work with learned binary routing for multi-task wireless optimization, but the routing component is not isolated from sparse-mask and retraining effects. read the letter →

arxiv 2502.10027 v1 pith:NGPPVRX7 submitted 2025-02-14 cs.LG

Nikos A. Mitsiou , Pavlos S. Bouzinis , Panagiotis G. Sarigiannidis , George K. Karagiannidis This is my paper · ORCID
classification cs.LG
keywords multi-tasklearningconditionalcomputationroutingDNNwirelessresourceallocationdeepforoptimizationsupervisedandunsupervisedvariabledimensionalitypower
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

Wireless networks face many resource-allocation problems that differ in the number of users or channels, in what is being optimized, and in the constraints. The paper proposes a single deep network that can handle all of these variants without retraining a new network for each one. The mechanism is conditional computation with routing: a small router network picks a subnetwork of a shared base network for each task by multiplying the base weights with binary masks. In experiments covering fourteen tasks, from convex delay minimization with labels to stochastic sum-capacity maximization without labels, the routed multi-task network nearly matches separately trained single-task networks, while naive parameter sharing and zero-padding fall clearly behind.

What carries the argument

The carrying object is the pair (bDNN, rDNN) with hard parameter sharing. The rDNN maps a one-hot task index $z_i$ to binary matrices $H^{l,i}$ for each hidden layer $l$; these are multiplied element-wise with the base weights, giving task-specific parameters $\Theta_i = \{ W^1[1:N_i,:], b^1, \{ H^{l,i} \odot W^l, b^l \}_{l=2}^L, W^{L+1}[1:N_i,:], b^L \}$ (Eq. 14), so each task flows through a distinct subnetwork of the same total parameter set. The hard threshold $y^{r,i} = \mathrm{Sign}(\mathrm{ReLU}(y^{r,i}(\Phi,z_i)-0.5))$ (Eq. 15) converts soft outputs into binary masks. Training first updates $\Omega=\{\Theta,\Phi\}$ jointly, including primal-dual steps for unsupervised constraints, then freezes the router and retrains the base network.

What would settle it

Re-run the fourteen-task experiment with the router allowed to see the input realization as well as the task index, and compare against the frozen per-task masks; if per-input routing yields a material gain in objective value or feasibility, the paper's fixed-path-per-task premise is falsified.

Watch

Extended reading notes

Core claim

The paper claims that hard parameter sharing with learned routing is enough to make multi-task learning work for heterogeneous optimization problems. During joint training, the router outputs soft masks; after training, these are thresholded at 0.5 and frozen, and each task runs through its own masked subnetwork of the shared base network, after which the base network is retrained. On the fourteen tested tasks, the proposed scheme stays close to the optimal single-task DNN, while naive multi-task and zero-padding baselines are substantially worse. The same framework supports both supervised tasks, where optimal labels are available, and unsupervised tasks, where a primal-dual Lagrangian loss enforces constraints.

Load-bearing premise

The framework assumes a single frozen subnetwork, found during joint training and fixed by a 0.5 threshold, remains the right computation path for every instance of a task, even though the router never sees the task's actual data.

Editorial extensions

If this is right

  • A single trained allocator can serve network configurations of different sizes, objectives, and constraints without changing architecture or retraining from scratch.
  • Because routing uses the same parameter set for all tasks, the model size stays comparable to one single-task DNN rather than growing with the number of tasks.
  • Switching between tasks at inference time reduces to changing the one-hot task index, which is cheap and compatible with real-time operation.
  • The framework applies to both supervised problems with known optimal solutions and unsupervised non-convex or stochastic problems where optimal labels cannot be generated.
  • Training complexity is roughly a constant factor larger than training K separate DNNs, since the router overhead scales with the number of tasks but is independent of the base network's per-task cost.

Reading between the lines

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

  • A natural extension the paper leaves implicit is per-instance routing: giving the router the channel realization as well as the task index would test whether the optimal computation path ever varies within a single task, which the current frozen-mask design assumes it does not.
  • The framework's behavior on unseen tasks is untested; a concrete next experiment would be to take a trained bDNN and either evaluate a new task index directly or fine-tune the router on a new objective, to see whether routing transfers or must be relearned.
  • Because the final binary masks define sparse task-specific subnetworks, the method connects naturally to pruning and architecture search: the masks could be analyzed as a learned allocation of parameters across tasks, possibly yielding interpretable per-task model sizes.
  • In an Open RAN setting, the authors' consolidation motivation could be tested end-to-end by deploying one multi-task model in place of several single-task xApps and measuring lifecycle overhead against the small performance gap reported here.
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.

Referee Report

4 major / 4 minor

Summary. The paper proposes a multi-task learning (MTL) architecture for solving wireless resource allocation problems of varying dimensionality, objective, and constraints with a single DNN. The architecture consists of a base DNN (bDNN) shared across tasks and a routing DNN (rDNN) that outputs binary masks, which are multiplied element-wise with the bDNN weights to define a task-specific subnetwork (Eqs. (13)-(15)). Tasks are defined as different network sizes N in {5,...,20} for an FDMA delay-minimization problem (supervised) and an average sum capacity maximization problem with average power constraints (unsupervised), giving fourteen tasks. Algorithm 1 trains the two networks jointly with soft masks, thresholds the masks at 0.5, and then retrains the bDNN with fixed masks. Numerical results compare the proposed scheme with single-task DNNs, zero-padding, and a naive multi-task baseline without routing, and report that the proposed method performs close to single-task DNNs while the baselines degrade.

Significance. If the empirical findings are reliable, the framework is a useful step toward DNN-based optimization that can adapt to variable problem dimensions and conflicting objectives without retraining from scratch, and the complexity analysis suggests training cost grows only by a constant factor relative to independent single-task DNNs. The paper is honest about the main design choice: routing depends only on the one-hot task indicator, so each task is assigned one fixed subnetwork for all channel realizations. The strengths include a clear formulation of the multi-task mapping, a broad fourteen-task testbed spanning supervised and unsupervised settings, and an explicit two-phase training procedure. However, the contribution hinges on whether the measured gains are caused by the learned routing decisions rather than by the sparse subnetwork structure and the retraining schedule; the current experiments do not isolate that mechanism, and the tables report point estimates without variance or significance tests.

major comments (4)
  1. [Section V, Tables II-III and Algorithm 1] To support the claim that the rDNN's learned routing is responsible for the observed gains, the experimental comparison must include fixed random masks with matched sparsity and a naive baseline that undergoes the same two-phase retraining loop. As written, the naive multi-task DNN is trained once without the threshold-and-retrain procedure, so the retraining phase and the per-task sparse structure are confounded with the routing objective. Without such an ablation, the results support the overall MTL architecture but do not establish that the specific routing decisions learned by the rDNN drive the improvement.
  2. [Section V, Tables II-III] The tables report only point estimates, with no error bars, multiple training seeds, or statistical significance tests. The 'near-single-task' claim rests on small absolute differences, e.g., 0.0152 vs. 0.0171 for the SL task at N=20 and 3.883 vs. 3.863 for the UL task at N=20, which may be within run-to-run variation. Please report the mean and standard deviation over several independent training runs and a paired comparison across tasks, so the reader can assess whether the proposed method is statistically indistinguishable from the single-task DNN.
  3. [Section V, parameter-fair comparison] The claimed parameter-fair comparison between the single-task DNN and the bDNN is not fully specified. For tasks with N_i < max{N_1,...,N_K}, the single-task DNN's input and output layers have dimension N_i, while the bDNN uses d1 = dL = max{N_1,...,N_K} (Eq. (12)) unless the hidden widths are adjusted separately. Thus the parameter counts will generally differ. Please state the exact architectures and parameter counts for every scheme, including zero-padding and the naive multi-task baseline, so the fairness of the comparison can be verified.
  4. [Section III-B, Eq. (15)] The hard thresholding step Sign(ReLU(yr,i - 0.5)) can in principle destroy the information in the jointly trained soft masks, and the paper provides no quantitative sensitivity analysis. The statement that retraining the bDNN 'slightly improves' performance is not accompanied by a comparison of final task performance before and after thresholding, nor by a sweep over the threshold value. Please add an ablation that reports the distribution of learned soft-mask values and the effect of the threshold on the final objective values, since the thresholding step is an essential part of the central mechanism.
minor comments (4)
  1. [Section V, Fig. 5 and surrounding text] The text refers to 'Table 3' and 'Table 4' when discussing the results that appear in Tables II and III; please correct the cross-references.
  2. [Fig. 6 caption] The caption contains the typo 'Constrain Violation' instead of 'Constraint Violation'.
  3. [Algorithm 1, line 10] The Sign function used in yr,i <- Sign(ReLU(yr,i - 0.5)) is not defined at zero; for completeness, please state that Sign(0) = 0.
  4. [Abstract and Section III-B] The abstract states that the rDNN's binary vector is multiplied with all bDNN weights, while Eq. (14) explicitly excludes the first and last layers; please adjust the wording for consistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central MTL/rDNN result is an empirical benchmark comparison, not a derivation from fitted constants or self-cited theorems.

full rationale

The paper's central claim is that the proposed bDNN+rDNN architecture jointly solves 14 optimization tasks with near-single-task performance. This is established by Tables II and III and the training curves against external or independently defined baselines: the single-task DNN of [14], zero-padding [20]-[22], and a naive all-ones-router multi-task network. No target quantity used in the evaluation is defined as a fitted parameter of the model, and no performance number is forced by construction: the rDNN masks are learned from the joint loss (Eq. 16) and then thresholded (Eq. 15), after which the bDNN is retrained (Algorithm 1). The evaluation metrics (delay MSE and average sum capacity) come from the original optimization problems (P3) and (P4), not from the model's own outputs. The only self-citation of note, [1], is used to identify the naive multi-task baseline, whose output is fixed to one everywhere rather than learned; it is a benchmark, not evidence for the central claim. The paper also cites [14] for the theoretical optimality of the single-task DNN, but that is an external result and is used as an upper/lower performance bound rather than as an assumption entailing the proposed method. There is no self-definitional step, no fitted input renamed as prediction, no imported uniqueness theorem, and no ansatz smuggled in via citation. The absence of a random-mask ablation is a possible confounding issue for the interpretation of the learned routing, but it is an experimental-control concern, not a circularity in the derivation chain.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central performance claim rests on hand-chosen hyperparameters (task weights, routing threshold, tanh steepness) and on assumptions that the primal-dual training converges and that task-index-only routing is sufficient. No new physical entities are introduced.

free parameters (5)
  • gamma = 5
    Steepness of tanh in the rDNN activation ReLU(tanh(gamma*x)); chosen by hand and controls the soft-to-binary approximation.
  • routing threshold = 0.5
    In Eq. (15), outputs below 0.5 are forced to 0 and above to 1; this hard threshold converts soft masks to binary and is a design choice not derived from the problem.
  • task loss weights beta_i = 1/K
    Uniform task weighting in Eq. (16). The paper does not tune or justify non-uniform weights, and the comparison depends on these weights.
  • rDNN hidden width T = 20
    Selected in Table II; affects routing capacity and complexity.
  • rDNN weight initialization mean and variance = mu=0.1, sigma=0.001
    Chosen offline by running a few simulations on the rDNN (Section III-B1), i.e., hand-fitted to make training work.
assumptions (4)
  • domain assumption The objective and constraint functions are differentiable; for the SL case the constraint functions are convex so optimal labels exist.
    Used in Sections II-A and II-B to generate datasets and to justify the primal-dual updates.
  • domain assumption The primal-dual DNN training converges to a near-optimal solution for the considered problems, following the single-task analysis of [14].
    The paper does not prove convergence for the multi-task setting and relies on prior single-task results.
  • ad hoc to paper A fixed per-task subnetwork, selected only by the one-hot task index, can approximate the optimal mapping for all channel realizations of that task.
    The rDNN input is the task index only; if intra-task instance-dependent routing were needed, the architecture would be insufficient.
  • standard math The constraint set X can be folded into the inequality constraints without loss of generality.
    Stated in Section II-B before discarding the explicit projection onto X.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Heterogeneous Resource Allocation with Multi-task Learning for Wireless Networks." pith.science (2026). https://pith.science/paper/NGPPVRX7

@misc{pith2026250210027,
  author       = {Pith},
  title        = {Pith review of: Heterogeneous Resource Allocation with Multi-task Learning for Wireless Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NGPPVRX7}},
  note         = {Machine review of arXiv:2502.10027}
}
read the original abstract

The optimal solution to an optimization problem depends on the problem's objective function, constraints, and size. While deep neural networks (DNNs) have proven effective in solving optimization problems, changes in the problem's size, objectives, or constraints often require adjustments to the DNN architecture to maintain effectiveness, or even retraining a new DNN from scratch. Given the dynamic nature of wireless networks, which involve multiple and diverse objectives that can have conflicting requirements and constraints, we propose a multi-task learning (MTL) framework to enable a single DNN to jointly solve a range of diverse optimization problems. In this framework, optimization problems with varying dimensionality values, objectives, and constraints are treated as distinct tasks. To jointly address these tasks, we propose a conditional computation-based MTL approach with routing. The multi-task DNN consists of two components, the base DNN (bDNN), which is the single DNN used to extract the solutions for all considered optimization problems, and the routing DNN (rDNN), which manages which nodes and layers of the bDNN to be used during the forward propagation of each task. The output of the rDNN is a binary vector which is multiplied with all bDNN's weights during the forward propagation, creating a unique computational path through the bDNN for each task. This setup allows the tasks to either share parameters or use independent ones, with the decision controlled by the rDNN. The proposed framework supports both supervised and unsupervised learning scenarios. Numerical results demonstrate the efficiency of the proposed MTL approach in solving diverse optimization problems. In contrast, benchmark DNNs lacking the rDNN mechanism were unable to achieve similar levels of performance, highlighting the effectiveness of the proposed architecture.

Figures

Figures reproduced from arXiv: 2502.10027 by the authors.

Figure 1
Figure 1. A modular sharing example for MTL. 2) Modular Sharing for MTL: Let us consider the compu￾tational graph of [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. The proposed multi-task DNN architecture. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. The proposed rDNN architecture. another task. Similarly, some weights can be used during the forward propagation of multiple tasks, or during the forward propagation of none task. 2) The training procedure of the multi-task DNN: Follow￾ing the analysis of the previous subsections we define the loss function of the multi-task DNN, including both the SL-based and the UL-based tasks, as follows L(Ω) = X i∈KS βi " 1 Bi … view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: The ReLU(tanh(γx)) and its derivative with respect to γ. + X Ni n=1 λ t−1 n,i 1 Bi X u∈Bi ∇Ωfn,i Ä y L+1 i (Θi) ; a u i ä !# . (19) λ t n,i = λ t−1 n,i + η t [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 6
Figure 6. Figure 6: The UL-based multi-task performance for N = 5. mitigate the interference between different tasks. The ZP scheme, for N = 20 obtains a value for the average sum capacity which is near to the value of the single-task DNN, but with increased constraint violation. Again, t…
Figure 5
Figure 5. Figure 5: The SL-based multi-task performance. single-task scheme. Finally, in [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 7
Figure 7. Figure 7: The UL-based multi-task performance for N = 12. training parameters directly influencing each task during both forward and backward propagation. Simulation results vali￾dated the efficiency of the proposed multi-task scheme. Future directions will aim to integrate the …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 34 canonical work pages

  1. [1]

    Multi-Task Learning for Resource Allocation in Wireless Networks of Dynamic Dimensionality,

    N. A. Mitsiou, P. S. Bouzinis, P. D. Diamantoulakis, P. G. Sarigiannidis, and G. K. Karagiannidis, “Multi-Task Learning for Resource Allocation in Wireless Networks of Dynamic Dimensionality,” in IEEE Interna- tional Symposium on Personal, Indoor and Mobile Radio Communica- tions (PIMRC) , Valencia, Spain, Sep. 2-5 2024

    work page 2024

  2. [2]

    Wireless Networks Design in the Era of Deep Learning: Model-Based, AI-Based, or Both?

    A. Zappone, M. Di Renzo, and M. Debbah, “Wireless Networks Design in the Era of Deep Learning: Model-Based, AI-Based, or Both?” IEEE Trans. Commun., vol. 67, no. 10, pp. 7331–7376, 2019

    work page 2019

  3. [3]

    Multiple access in the era of distributed computing and edge intelligence,

    N. G. Evgenidis, N. A. Mitsiou, V . I. Koutsioumpa, S. A. Tegos, P. D. Diamantoulakis, and G. K. Karagiannidis, “Multiple access in the era of distributed computing and edge intelligence,” Proc. IEEE, pp. 1–30, 2024

    work page 2024

  4. [4]

    Model-Based Deep Learning,

    N. Shlezinger, J. Whang, Y . C. Eldar, and A. G. Dimakis, “Model-Based Deep Learning,” Proc. IEEE, vol. 111, no. 5, pp. 465–499, 2023

    work page 2023

  5. [5]

    Predictive quality of service: The next frontier for fully autonomous systems,

    M. Boban, M. Giordani, and M. Zorzi, “Predictive quality of service: The next frontier for fully autonomous systems,” IEEE Network, vol. 35, no. 6, pp. 104–110, 2021

    work page 2021

  6. [6]

    From Multilayer Perceptron to GPT: A Reflection on Deep Learning Research for Wireless Physical Layer,

    M. Akrout, A. Mezghani, E. Hossain, F. Bellili, and R. W. Heath, “From Multilayer Perceptron to GPT: A Reflection on Deep Learning Research for Wireless Physical Layer,” arXiv preprint arXiv:2307.07359 , 2023. 0 1000 2000 3000 4000 50003 3.5 4 4.5 Number of Iterations Average Sum Capacity single-task DNN zero-padding naive multi-task DNN proposed multi-ta...

    arXiv 2023

  7. [7]

    Multi-task learning with deep neural networks: A survey,

    M. Crawshaw, “Multi-task learning with deep neural networks: A survey,” arXiv preprint arXiv:2009.09796 , 2020

    arXiv 2009

  8. [8]

    Routing networks: Adaptive selection of non-linear functions for multi-task learning,

    C. Rosenbaum, T. Klinger, and M. Riemer, “Routing networks: Adaptive selection of non-linear functions for multi-task learning,” arXiv preprint arXiv:1711.01239, 2017

    arXiv 2017

Show all 44 references
  1. [9]

    Multi-task reinforcement learning with soft modularization,

    R. Yang, H. Xu, Y . Wu, and X. Wang, “Multi-task reinforcement learning with soft modularization,” Advances in Neural Information Processing Systems, vol. 33, pp. 4767–4777, 2020

    2020

  2. [10]

    Once-for-all: Train one network and specialize it for efficient deployment,

    H. Cai, C. Gan, T. Wang, Z. Zhang, and S. Han, “Once-for-all: Train one network and specialize it for efficient deployment,” arXiv preprint arXiv:1908.09791, 2019

    1908 arXiv

  3. [11]

    Dynamic Neural Networks: A Survey,

    Y . Han, G. Huang, S. Song, L. Yang, H. Wang, and Y . Wang, “Dynamic Neural Networks: A Survey,” IEEE Trans. Pattern Anal. Mach. Intell. , vol. 44, no. 11, pp. 7436–7456, 2022

    2022

  4. [12]

    Slimmable neural networks,

    J. Yu, L. Yang, N. Xu, J. Yang, and T. Huang, “Slimmable neural networks,” arXiv preprint arXiv:1812.08928 , 2018

    2018 arXiv

  5. [13]

    Learning to optimize: Training deep neural networks for wireless resource management,

    H. Sun, X. Chen, Q. Shi, M. Hong, X. Fu, and N. D. Sidiropoulos, “Learning to optimize: Training deep neural networks for wireless resource management,” in 2017 IEEE 18th International Workshop on Signal Processing Advances in Wireless Communications (SPA WC) , 2017, pp. 1–6

    2017

  6. [14]

    Deep Learning for Distributed Optimization: Applications to Wireless Resource Management,

    H. Lee, S. H. Lee, and T. Q. S. Quek, “Deep Learning for Distributed Optimization: Applications to Wireless Resource Management,” IEEE J. Sel. Areas Commun. , vol. 37, no. 10, pp. 2251–2266, 2019

    2019

  7. [15]

    Intelligent Resource Allocation in Wireless Communications Systems,

    W. Lee, O. Jo, and M. Kim, “Intelligent Resource Allocation in Wireless Communications Systems,” IEEE Commun. Mag. , vol. 58, no. 1, pp. 100–105, 2020

    2020

  8. [16]

    Online Deep Neural Network for Optimization in Wireless Communications,

    J. Gao, C. Zhong, G. Y . Li, and Z. Zhang, “Online Deep Neural Network for Optimization in Wireless Communications,” IEEE Wirel. Commun. Lett., vol. 11, no. 5, pp. 933–937, 2022

    2022

  9. [17]

    Intelligent Over-the-Air Computing Environ- ment,

    P. S. Bouzinis, N. A. Mitsiou, P. D. Diamantoulakis, D. Tyrovolas, 13 and G. K. Karagiannidis, “Intelligent Over-the-Air Computing Environ- ment,” IEEE Wirel. Commun. Lett. , vol. 12, no. 1, pp. 134–137, 2023

    2023

  10. [18]

    Meta-Learning for Wireless Communications: A Survey and a Comparison to GNNs,

    B. Zhao, J. Wu, Y . Ma, and C. Yang, “Meta-Learning for Wireless Communications: A Survey and a Comparison to GNNs,” IEEE open j. Commun. Soc. , vol. 5, pp. 1987–2015, 2024

    1987

  11. [19]

    Modular Meta-Learning for Power Control via Random Edge Graph Neural Networks,

    I. Nikoloska and O. Simeone, “Modular Meta-Learning for Power Control via Random Edge Graph Neural Networks,” IEEE Trans. Wirel. Commun., vol. 22, no. 1, pp. 457–470, 2023

    2023

  12. [20]

    LoRa- RL: Deep Reinforcement Learning for Resource Management in Hybrid Energy LoRa Wireless Networks,

    R. Hamdi, E. Baccour, A. Erbad, M. Qaraqe, and M. Hamdi, “LoRa- RL: Deep Reinforcement Learning for Resource Management in Hybrid Energy LoRa Wireless Networks,” IEEE Internet Things J. , vol. 9, no. 9, pp. 6458–6476, 2022

    2022

  13. [21]

    Resource Management in Wireless Networks via Multi-Agent Deep Reinforce- ment Learning,

    N. Naderializadeh, J. J. Sydir, M. Simsek, and H. Nikopour, “Resource Management in Wireless Networks via Multi-Agent Deep Reinforce- ment Learning,” IEEE Trans. Wirel. Commun. , vol. 20, no. 6, pp. 3507– 3523, 2021

    2021

  14. [22]

    A Deep Learning based Resource Allocation Algorithm for Variable Dimensions in D2D-Enabled Cellular Networks,

    E. Pei and G. Yang, “A Deep Learning based Resource Allocation Algorithm for Variable Dimensions in D2D-Enabled Cellular Networks,” in 2020 IEEE/CIC International Conference on Communications in China (ICCC) , 2020, pp. 277–282

    2020

  15. [23]

    An Overview on the Application of Graph Neural Networks in Wireless Networks,

    S. He, S. Xiong, Y . Ou, J. Zhang, J. Wang, Y . Huang, and Y . Zhang, “An Overview on the Application of Graph Neural Networks in Wireless Networks,” IEEE Open Journal of the Communications Society , vol. 2, pp. 2547–2565, 2021

    2021

  16. [24]

    Optimal Wireless Resource Allocation With Random Edge Graph Neural Networks,

    M. Eisen and A. Ribeiro, “Optimal Wireless Resource Allocation With Random Edge Graph Neural Networks,” IEEE Transactions on Signal Processing, vol. 68, pp. 2977–2991, 2020

    2020

  17. [25]

    Graph Neural Networks for Scalable Radio Resource Management: Architecture Design and Theo- retical Analysis,

    Y . Shen, Y . Shi, J. Zhang, and K. B. Letaief, “Graph Neural Networks for Scalable Radio Resource Management: Architecture Design and Theo- retical Analysis,” IEEE Journal on Selected Areas in Communications , vol. 39, no. 1, pp. 101–115, 2021

    2021

  18. [26]

    Learning Decentralized Wireless Resource Allocations With Graph Neural Networks,

    Z. Wang, M. Eisen, and A. Ribeiro, “Learning Decentralized Wireless Resource Allocations With Graph Neural Networks,” IEEE Transactions on Signal Processing , vol. 70, pp. 1850–1863, 2022

    2022

  19. [27]

    Learning Resilient Radio Resource Management Policies With Graph Neural Networks,

    N. NaderiAlizadeh, M. Eisen, and A. Ribeiro, “Learning Resilient Radio Resource Management Policies With Graph Neural Networks,” IEEE Transactions on Signal Processing , vol. 71, pp. 995–1009, 2023

    2023

  20. [28]

    Learning power allocation for multi-cell-multi-user systems with heterogeneous graph neural networks,

    J. Guo and C. Yang, “Learning power allocation for multi-cell-multi-user systems with heterogeneous graph neural networks,” IEEE Transactions on Wireless Communications , vol. 21, no. 2, pp. 884–897, 2022

    2022

  21. [29]

    Multi-task self-supervised graph neural networks enable stronger task generalization,

    M. Ju, T. Zhao, Q. Wen, W. Yu, N. Shah, Y . Ye, and C. Zhang, “Multi-task self-supervised graph neural networks enable stronger task generalization,” arXiv preprint arXiv:2210.02016 , 2022

    2022 arXiv

  22. [30]

    All in One: Multi-Task Prompting for Graph Neural Networks,

    X. Sun, H. Cheng, J. Li, B. Liu, and J. Guan, “All in One: Multi-Task Prompting for Graph Neural Networks,” in Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , ser. KDD ’23. New York, NY , USA: Association for Computing Machinery, 2023, p....

    2023

  23. [31]

    Evolutionary Multitasking With Dynamic Resource Allocating Strategy,

    M. Gong, Z. Tang, H. Li, and J. Zhang, “Evolutionary Multitasking With Dynamic Resource Allocating Strategy,” IEEE Transactions on Evolutionary Computation , vol. 23, no. 5, pp. 858–869, 2019

    2019

  24. [32]

    Towards Generalized Resource Allocation on Evolutionary Multitasking for Multi-Objective Optimization,

    T. Wei and J. Zhong, “Towards Generalized Resource Allocation on Evolutionary Multitasking for Multi-Objective Optimization,” IEEE Computational Intelligence Magazine , vol. 16, no. 4, pp. 20–37, 2021

    2021

  25. [33]

    [a review on evolutionary multitask optimization: Trends and challenges],

    T. Wei, S. Wang, J. Zhong, D. Liu, and J. Zhang, “[a review on evolutionary multitask optimization: Trends and challenges],” IEEE Transactions on Evolutionary Computation , vol. 26, no. 5, pp. 941–960, 2022

    2022

  26. [34]

    Adaptive Auxiliary Task Selection for Multitasking-Assisted Constrained Multi-Objective Optimization [Fea- ture],

    F. Ming, W. Gong, and L. Gao, “Adaptive Auxiliary Task Selection for Multitasking-Assisted Constrained Multi-Objective Optimization [Fea- ture],” IEEE Computational Intelligence Magazine , vol. 18, no. 2, pp. 18–30, 2023

    2023

  27. [35]

    Constrained Multiobjective Optimization via Multitasking and Knowledge Transfer,

    F. Ming, W. Gong, L. Wang, and L. Gao, “Constrained Multiobjective Optimization via Multitasking and Knowledge Transfer,” IEEE Trans- actions on Evolutionary Computation , vol. 28, no. 1, pp. 77–89, 2024

    2024

  28. [36]

    Un- derstanding O-RAN: Architecture, Interfaces, Algorithms, Security, and Research Challenges,

    M. Polese, L. Bonati, S. D’Oro, S. Basagni, and T. Melodia, “Un- derstanding O-RAN: Architecture, Interfaces, Algorithms, Security, and Research Challenges,” IEEE Commun. Surv. Tutor ., vol. 25, no. 2, pp. 1376–1411, 2023

    2023

  29. [37]

    Learning to detect,

    N. Samuel, T. Diskin, and A. Wiesel, “Learning to detect,” IEEE Trans. Signal Process., vol. 67, no. 10, pp. 2554–2564, 2019

    2019

  30. [38]

    Optnet: Differentiable optimization as a layer in neural networks,

    B. Amos and J. Z. Kolter, “Optnet: Differentiable optimization as a layer in neural networks,” in International Conference on Machine Learning . PMLR, 2017, pp. 136–145

    2017

  31. [39]

    Boyd and L

    S. Boyd and L. Vandenberghe, Convex optimization . Cambridge university press, 2004

    2004

  32. [40]

    Confi- dence Aware Deep Learning Driven Wireless Resource Allocation in Shared Spectrum Bands,

    C. Ganewattha, Z. Khan, M. Latva-Aho, and J. J. Lehtom ¨aki, “Confi- dence Aware Deep Learning Driven Wireless Resource Allocation in Shared Spectrum Bands,” IEEE Access , vol. 10, pp. 34 945–34 959, 2022

    2022

  33. [41]

    Proactive Resource Management for LTE in Unlicensed Spectrum: A Deep Learning Perspective,

    U. Challita, L. Dong, and W. Saad, “Proactive Resource Management for LTE in Unlicensed Spectrum: A Deep Learning Perspective,” IEEE Trans. Wirel. Commun. , vol. 17, no. 7, pp. 4674–4689, 2018

    2018

  34. [42]

    6g wireless networks: Vision, requirements, architecture, and key technologies,

    Z. Zhang, Y . Xiao, Z. Ma, M. Xiao, Z. Ding, X. Lei, G. K. Karagiannidis, and P. Fan, “6g wireless networks: Vision, requirements, architecture, and key technologies,” IEEE V eh. Technol. Mag. , vol. 14, no. 3, pp. 28–41, 2019

    2019

  35. [43]

    A review of Wi-Fi 6: The revolution of 6th generation Wi-Fi technology,

    A. S. George and A. H. George, “A review of Wi-Fi 6: The revolution of 6th generation Wi-Fi technology,” Res. Inventy Int. J. Eng. Sci. , vol. 10, pp. 56–65, 2020

    2020

  36. [44]

    IEEE 802.11 be Wi-Fi 7: New challenges and opportunities,

    C. Deng, X. Fang, X. Han, X. Wang, L. Yan, R. He, Y . Long, and Y . Guo, “IEEE 802.11 be Wi-Fi 7: New challenges and opportunities,” IEEE Commun. Surv. Tutor ., vol. 22, no. 4, pp. 2136–2166, 2020

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.