Pith. sign in

REVIEW 3 major objections 4 minor 29 references

COMIX: Generalized Conflict Management in O-RAN xApps -- Architecture, Workflow, and a Power Control case

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read COMIX, an O-RAN conflict management scheme with a network digital twin, resolves clashes between throughput-maximizing and energy-efficiency xApps and cuts power consumption versus a conflict-free baseline.

desk verdict A coherent O-RAN conflict-management architecture whose headline energy savings are not established: the baseline is effectively random, the policies almost always pick EE, and the NDT is the same simulator used for training. read the letter →

arxiv 2501.14619 v1 pith:PIHAFITH submitted 2025-01-24 cs.NI

Anastasios Giannopoulos , Sotirios Spantideas , Levis George , Kalafatelis Alexandros , Panagiotis Trakadas This is my paper · ORCID
classification cs.NI
keywords O-RANconflictmanagementxAppdeepreinforcementlearningpowercontrolnetworkdigitaltwinenergyefficiencyNear-RTRIC
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

The paper proposes COMIX, a conflict management scheme for O-RAN that detects and resolves clashes between xApps running on the Near-Real-Time RIC. Its demonstration uses two deep reinforcement learning xApps that both tune radio-unit transmit power: one maximizes user data rate, the other maximizes system energy efficiency. Because both write the same control parameter, their actions directly conflict whenever one raises power and the other lowers it. COMIX resolves the conflict by evaluating each candidate action in a Network Digital Twin, under operator-chosen resolution policies, before applying an action to the live network. The authors show that this arbitration consistently lowers total power consumption compared to a conflict-free baseline that applies whichever xApp action arrives last, while preserving data rate.

What carries the argument

The central mechanism is the Conflict Mitigation Framework pipeline inside the Near-RT RIC: a Conflict Detector that builds clusters of control parameters per KPI from an association matrix $A$, so it can flag a direct conflict when two xApps share a control parameter and an indirect conflict when their parameters influence the same KPI; a KPI Notifier that later expands clusters when unexpected KPI degradation reveals implicit conflicts; and a Conflict Resolver that uses the Network Digital Twin to simulate each conflicting action and ranks the actions by the active resolution policy before the Action Taker applies the winner.

What would settle it

Run the DRM and EE xApps on a testbed or field deployment where the NDT is calibrated against live-channel measurements rather than the training simulator; if the CMF-resolved actions do not reduce total RU power consumption below the last-write baseline over a comparable validation run, the central claim fails.

Watch

Extended reading notes

Core claim

COMIX is a CMF-aligned conflict management framework whose conflict detector classifies xApp disagreements into direct, indirect, and implicit conflicts using control-parameter and KPI association data; its conflict resolver then scores each candidate action through a Network Digital Twin under one of five resolution policies (MaxTS, MinPS, EES, TVS, EEVS) and applies the best-scoring action. In validation over 200 episodes with DRM and EE xApps, the CMF-based scheme increases energy savings relative to the CMF-free baseline across all five policies, while data rate is preserved and only slightly exceeds the baseline under the throughput-maximizing policy.

Load-bearing premise

The validation assumes the Network Digital Twin used to rank candidate actions ranks them the same way the live network would, since the same simulator both trains the xApps and evaluates the conflicting actions.

Editorial extensions

If this is right

  • Operators can run throughput-focused and energy-focused xApps simultaneously without a hard-wired priority, because the resolver picks per-slot actions based on current network simulation rather than a static rule.
  • Power-conscious resolution policies (MinPS, EES, EEVS) deliver the largest energy savings, so operators targeting sustainability can choose among them without modifying the xApps.
  • The same CMF pipeline extends beyond power control to any control parameter shared by xApps, since detection relies only on parameter/KPI association data.
  • The NDT-based evaluation turns conflict resolution from a reactive last-write-wins process into a prospective one, avoiding degradation of the live network before actions are applied.

Reading between the lines

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

  • If the NDT's fidelity degrades in a live deployment, the resolver's action ranking may diverge from the real network's response, so the scheme's energy-savings guarantee depends on keeping the twin calibrated.
  • The framework could be extended to more than two xApps by evaluating groups of non-conflicting action combinations, which the current pairwise demonstration does not explore.
  • A natural test is to wire COMIX to an open-source O-RAN emulator with a channel model different from the training simulator; if the energy savings vanish, the result would isolate NDT accuracy as the load-bearing factor.
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

3 major / 4 minor

Summary. The paper proposes COMIX, a conflict management framework for O-RAN Near-RT RIC that combines a standardized Conflict Mitigation Framework (CMF) with a Network Digital Twin (NDT) to detect and resolve conflicts between xApps. Two deep reinforcement learning (DRL) xApps, one maximizing system data rate (DRM) and one maximizing energy efficiency (EE) under relaxed QoS constraints, are designed for RU power control. Five resolution policies (MaxTS, MinPS, EES, TVS, EEVS) are evaluated against a CMF-free baseline where the last-arriving xApp action is applied. The paper reports that all policies yield large energy savings relative to the baseline and that NDT-based evaluation is central to the scheme, while acknowledging that the NDT is the same software package used for training.

Significance. If the validation were convincing, COMIX would be a useful architectural contribution to O-RAN conflict management, and the comparison of five operator-defined resolution policies is a reasonable way to explore the design space. The paper provides a clear formalization of conflict detection via CP/KPI association graphs, a detailed workflow, and a simulation-based case study. However, the current experimental design does not isolate the effect of the conflict management mechanism from the simple preference for the low-power EE xApp, and the absence of statistical uncertainty reporting further weakens the quantitative claims. The framework may still be valuable, but the evidence presented in this version does not substantiate the central claim that COMIX itself yields the reported energy savings.

major comments (3)
  1. [Section IV-B, Figs. 6 and 7] The central claim that COMIX 'considerably increases the energy savings of the system' is not supported by the presented comparison. The CMF-free baseline is described as 'effectively resulting in a random selection between the DRM and EE xApps,' while Fig. 7 shows that all five resolution policies select the EE xApp's action in the vast majority of time steps. The comparison therefore reduces to a policy that nearly always applies the low-power EE action versus a baseline that applies EE roughly half the time. A necessary control is a scheme that unconditionally applies the EE xApp's action at every step; without it, the results only demonstrate that prioritizing the EE xApp reduces power consumption, not that the CMF's conflict detection, NDT evaluation, or policy scoring is responsible for the savings.
  2. [Section IV-B, second paragraph] The paper states that 'the NDT environment considered for the proactive action assessment is the same software package used to train both xApps.' This creates an in-sample evaluation loop: the resolver's action scores are produced by the exact simulator in which the DRL agents were trained, so the experiment does not measure the NDT's ability to predict a live network or even a held-out environment. The claim in Section II-A6 that NDT 'improves decision accuracy pre-action' is therefore not validated. The manuscript's own Section V-B concedes that 'elaborating on how NDT could continuously provide accurate representations of the real O-RAN system remains of paramount importance,' which is appropriate, but the current results should be framed accordingly.
  3. [Section IV-B, Figs. 6 and 7] All quantitative performance claims rest on averages over 200 validation episodes without error bars, confidence intervals, or significance tests. In particular, the statement that MinPS, EES, and EEVS 'showed equally optimal performance' in power savings cannot be verified, and the magnitude of the claimed 'considerable' improvement over the baseline has no quantified uncertainty. At minimum, the standard deviation or interquartile range across episodes should be reported, and ideally a statistical test (e.g., paired bootstrap) should accompany comparisons among policies.
minor comments (4)
  1. [Abstract] The phrase 'maximizes the data rare across UEs' contains a typo; it should be 'data rate.'
  2. [Section III-C] The word 'framewrok' appears in the first paragraph; it should be 'framework.'
  3. [Figures 5 and 6] The captions and axis labels in the provided manuscript contain corrupted text (e.g., '/uni00000013/…' sequences), making the figures difficult to read. The production version should be checked for font-embedding issues.
  4. [Section IV-B, final paragraph] The unit 'Mpbs/Watt' should be 'Mbps/Watt.'

Circularity Check

1 steps flagged · score 6.0 of 10

Energy-savings claim reduces to the power-minimizing definition of the resolution policies, not to conflict-management value.

  1. self definitional [Section IV-B, Eqs. (9) and (10), Figs. 6b and 7]
    "Minimum Power-based Selection (MinPS): This policy resolves the direct conflicts by selecting the power vector that minimizes the total power consumed. Thus, the score of each action j is computed by: s2_j(t) = - sum_n sum_m p^j_n,m ... Importantly, the CMF considerably increases the energy savings of the system, as clearly illustrated in Fig. 6b."

    MinPS's selection score is the negative of total power, so the resolver's chosen action is, by definition, the lowest-power action. The headline energy-savings result then measures exactly the quantity that the policy is defined to optimize. Comparing this power-minimizing selector to a baseline that 'effectively results in a random selection between the DRM and EE xApps' guarantees lower power consumption whenever the EE action is lower-power, which Fig. 7 confirms by showing EE is chosen in the vast majority of steps. The same holds for EES and EEVS, whose scores include power or EE ratios.

full rationale

The COMIX architecture and conflict-detection graph (Section II) are self-contained and are not circular: the detector uses externally defined CP/KPI associations, and the resolution policies are stated independently of the results. However, the paper's most prominent quantitative result, the energy-savings improvement in Fig. 6b, is forced by construction for the power-aware policies. MinPS, EES, and EEVS score candidate actions using power consumption or energy efficiency, and the paper's own Fig. 7 shows that these policies choose the EE xApp's lower-power action almost always; the CMF-free baseline is described as 'effectively resulting in a random selection between the DRM and EE xApps.' Consequently, the comparison demonstrates only that selecting low-power actions reduces power relative to random selection. Additionally, Section IV-B states that 'the NDT environment considered for the proactive action assessment is the same software package used to train both xApps,' and Section V-B admits that 'elaborating on how NDT could continuously provide accurate representations of the real O-RAN system remains of paramount importance'; this is an in-sample evaluation loop that limits external validity, though it is not itself a definitional circularity. The framework retains independent architectural content, so the score is 6 rather than higher.

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

No new physical entities, forces, or conserved quantities are introduced. The framework introduces software components (Conflict Detector modules, resolution policies within CMF) that are architectural in nature. The main hidden burden is the assumed fidelity of the NDT and the hand-chosen SLA thresholds and throughput coefficient xi that shape the reported energy savings.

free parameters (4)
  • xi (throughput coefficient for EE xApp SLA relaxation) = 0.8
    Chosen by hand (Table III) to define the trade-off between data rate and energy saving in the EE xApp's constraint (C4'). It directly shapes the EE agent's learned policy and therefore what the resolver selects.
  • Power step Ps for EE xApp = 11 W
    Tuned in Section IV-A as the value giving the highest reward for the EE xApp. The power step defines the action set and strongly influences what actions the CMF can pick from.
  • Power step Ps for DRM xApp = 3 W
    Tuned in Section IV-A as optimal for DRM xApp. Again this defines the candidate action space resolved by CMF.
  • SLA thresholds cu and eu = 2 Mbps and 2 Mbps/W
    Set 'without loss of generality' in Section IV-B. These thresholds determine TVS/EEVS policy scores and hence which actions are chosen in the main energy-saving results.
assumptions (4)
  • domain assumption The NDT simulator represents the live network accurately enough to rank candidate actions correctly (Section IV-B, Fig. 6; Section V-B).
    All claims about NDT-based conflict resolution improving live-network decisions depend on fidelity of the simulator to the real RAN. The authors state this remains open in Section V-B.
  • domain assumption The CMF-free baseline (last-received action applies) is a representative no-conflict-management scenario (Section IV-B).
    The energy-savings comparison is against this baseline. Real deployments could use other arbitration (priority, voting, averaging), which might change the size of the reported benefit.
  • standard math Shannon capacity with complete frequency reuse, Eq. (1)-(4), models the multi-cell multi-channel network adequately for the conclusions.
    A standard idealization in wireless simulators; it ignores finite MCS tables, scheduling overhead, and control-channel effects.
  • domain assumption The association matrix A and fixed RB-UE association per time slot are given and static within a step, with dynamic reassociation happening only through the simulator's 'Dynamic association' procedure (Section III-C).
    The optimization constraints C3 and the user association procedure are assumed to be handled by the NDT; the paper does not specify the reassociation algorithm or its interaction with the power control actions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of COMIX: Generalized Conflict Management in O-RAN xApps -- Architecture, Workflow, and a Power Control case." pith.science (2026). https://pith.science/paper/PIHAFITH

@misc{pith2026250114619,
  author       = {Pith},
  title        = {Pith review of: COMIX: Generalized Conflict Management in O-RAN xApps -- Architecture, Workflow, and a Power Control case},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PIHAFITH}},
  note         = {Machine review of arXiv:2501.14619}
}
read the original abstract

Open Radio Access Network (O-RAN) is transforming the telecommunications landscape by enabling flexible, intelligent, and multi-vendor networks. Central to its architecture are xApps hosted on the Near-Real-Time RAN Intelligent Controller (Near-RT RIC), which optimize network functions in real time. However, the concurrent operation of multiple xApps with conflicting objectives can lead to suboptimal performance. This paper introduces a generalized Conflict Management scheme for Multi-Channel Power Control in O-RAN xApps (COMIX), designed to detect and resolve conflicts between xApps. To demonstrate COMIX, we focus on two Deep Reinforcement Learning (DRL)-based xApps for power control: one maximizes the data rare across UEs, and the other optimizes system-level energy efficiency. COMIX employs a standardized Conflict Mitigation Framework (CMF) for conflict detection and resolution and leverages the Network Digital Twin (NDT) to evaluate the impact of conflicting actions before applying them to the live network. We validate the framework using a realistic multi-channel power control scenario under various conflict resolution policies, demonstrating its effectiveness in balancing antagonistic objectives. Our results highlight significant network energy savings achieved through the conflict management scheme compared to baseline CMF-free methods.

Figures

Figures reproduced from arXiv: 2501.14619 by the authors.

Figure 1
Figure 1. Conflicts in O-RAN. (a) Conflict categories in O-RAN. (b) Architec￾tural view of conflict categories. (CMF) within the RIC architecture [8]. The CMF is tasked with detecting, classifying, and resolving conflicts among xApps and rApps to ensure optimal coordination and resource utilization. Conflicts in O-RAN can be categorized as shown in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. General O-RAN architecture considered for COMIX. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Generalized Direct/Indirect Conflict Detector. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: DRL cycles between xApps and O-RAN environment. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Learning curves of DRM (a-c) and EE (d-f) xApps for different learning rates (column 1), discount factors (column 2), and power steps (column 3). TABLE III SIMULATED ENVIRONMENT AND DRL CONFIGURATION Parameter Value Number Of RUs (N) 3 Number Of RBs (M) 8 Number Of UEs…
Figure 7
Figure 7. Figure 7: Frequency of occurrence for each xApp’s action (DRM and EE) across [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 6
Figure 6. Figure 6: Network Data Rate (a) and Total Power Consumption (b) resulted from different resolution policies (blue curves) after 200 validation episodes. Red curves reflect the same metrics without the presence of CMF. To demonstrate the performance of each resolution policy, we …
Figure 8
Figure 8. Figure 8: Sequence diagram and information flow in the COMIX scheme for multi-xApp inference and conflict management. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 27 canonical work pages

  1. [20]

    Learning to fulfill the user demands in 5g-enabled wireless networks through power allocation: A reinforcement learning approach,

    A. Giannopoulos, S. Spantideas, N. Nomikos, A. Kalafatelis, and P. Trakadas, “Learning to fulfill the user demands in 5g-enabled wireless networks through power allocation: A reinforcement learning approach,” in 2023 19th International Conference on the Design of Reliable Communication Networks (DRCN) . IEEE, 2023, pp. 1–7

    work page 2023

  2. [22]

    Deep reinforcement learning for energy-efficient multi- channel transmissions in 5g cognitive hetnets: Centralized, decentralized and transfer learning based solutions,

    A. Giannopoulos, S. Spantideas, N. Kapsalis, P. Karkazis, and P. Trakadas, “Deep reinforcement learning for energy-efficient multi- channel transmissions in 5g cognitive hetnets: Centralized, decentralized and transfer learning based solutions,” IEEE Access, vol. 9, pp. 129 358– 129 374, 2021

    work page 2021

  3. [1]

    Under- standing o-ran: Architecture, interfaces, algorithms, security, and re- search challenges,

    M. Polese, L. Bonati, S. D’oro, S. Basagni, and T. Melodia, “Under- standing o-ran: Architecture, interfaces, algorithms, security, and re- search challenges,” IEEE Communications Surveys & Tutorials, vol. 25, no. 2, pp. 1376–1411, 2023

    2023

  4. [2]

    Toward next generation open radio access networks: What o-ran can and cannot do!

    A. S. Abdalla, P. S. Upadhyaya, V . K. Shah, and V . Marojevic, “Toward next generation open radio access networks: What o-ran can and cannot do!” IEEE Network, vol. 36, no. 6, pp. 206–213, 2022

    work page 2022

  5. [3]

    Orchestran: Net- work automation through orchestrated intelligence in the open ran,

    S. D’Oro, L. Bonati, M. Polese, and T. Melodia, “Orchestran: Net- work automation through orchestrated intelligence in the open ran,” in IEEE INFOCOM 2022-IEEE Conference on Computer Communications. IEEE, 2022, pp. 270–279

    work page 2022

  6. [4]

    Supporting intelligence in disaggregated open radio access networks: Architectural principles, ai/ml workflow, and use cases,

    A. Giannopoulos, S. Spantideas, N. Kapsalis, P. Gkonis, L. Sarakis, C. Capsalis, M. Vecchio, and P. Trakadas, “Supporting intelligence in disaggregated open radio access networks: Architectural principles, ai/ml workflow, and use cases,” IEEE Access , vol. 10, pp. 39 580–39 595, 2022

    work page 2022

  7. [5]

    Intelligent mission critical services over beyond 5g networks: Control loop and proactive overload detection,

    S. Spantideas, A. Giannopoulos, M. A. Cambeiro, O. Trullols-Cruces, E. Atxutegi, and P. Trakadas, “Intelligent mission critical services over beyond 5g networks: Control loop and proactive overload detection,” in 2023 International Conference on Smart Applications, Communications and Networking (SmartNets) . IEEE, 2023, pp. 1–6

    work page 2023

  8. [6]

    Misconfig- uration in o-ran: Analysis of the impact of ai/ml,

    N. M. Yungaicela-Naula, V . Sharma, and S. Scott-Hayward, “Misconfig- uration in o-ran: Analysis of the impact of ai/ml,” Computer Networks, p. 110455, 2024

    2024

Show all 29 references
  1. [7]

    Conflict mitigation framework and conflict detection in o-ran near-rt ric,

    C. Adamczyk and A. Kliks, “Conflict mitigation framework and conflict detection in o-ran near-rt ric,” IEEE Communications Magazine, vol. 61, no. 12, pp. 199–205, 2023

    2023

  2. [8]

    O-ran working group 3 (near-real-time ran intelligent con- troller and e2 interface workgroup)(o-ran. wg3. ricarch-r003-v04. 00): Tech. rep

    O-RAN, “O-ran working group 3 (near-real-time ran intelligent con- troller and e2 interface workgroup)(o-ran. wg3. ricarch-r003-v04. 00): Tech. rep.” O-RAN Alliance, 2023. PREPRINT 14

    2023

  3. [9]

    Pre-emptive conflict detection architecture for o-ran service management and orchestration,

    J. Armstrong, E. Fallon, and S. Fallon, “Pre-emptive conflict detection architecture for o-ran service management and orchestration,” in 2024 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT) . IEEE, 2024, pp. 335–340

    2024

  4. [10]

    xapp-level conflict mitiga- tion in o-ran, a mobility driven energy saving case,

    A. Wadud, F. Golpayegani, and N. Afraz, “xapp-level conflict mitiga- tion in o-ran, a mobility driven energy saving case,” arXiv preprint arXiv:2411.03326, 2024

    2024 arXiv

  5. [11]

    Conflict management in the near-rt-ric of open ran: A game theoretic approach,

    A. Wadud, F. Golpayegani, and N. Afraz, “Conflict management in the near-rt-ric of open ran: A game theoretic approach,” in 2023 IEEE International Conferences on Internet of Things (iThings) and IEEE Green Computing & Communications (GreenCom) and IEEE Cyber, Physical & Socia...

    2023

  6. [12]

    Team learning-based re- source allocation for open radio access network (o-ran),

    H. Zhang, H. Zhou, and M. Erol-Kantarci, “Team learning-based re- source allocation for open radio access network (o-ran),” in ICC 2022- IEEE International Conference on Communications . IEEE, 2022, pp. 4938–4943

    2022

  7. [13]

    Towards closed-loop automation in 5g open ran: Coupling an open-source simulator with xapps,

    T. Karamplias, S. T. Spantideas, A. E. Giannopoulos, P. Gkonis, N. Kap- salis, and P. Trakadas, “Towards closed-loop automation in 5g open ran: Coupling an open-source simulator with xapps,” in 2022 Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G...

    2022

  8. [14]

    Digital twin for automated industrial optimization: Intelligent machine selection via process modelling,

    K. Skianis, A. Giannopoulos, A. Kalafatelis, and P. Trakadas, “Digital twin for automated industrial optimization: Intelligent machine selection via process modelling,” in 2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technolog...

    2023

  9. [15]

    O-ran: Disrupting the virtual- ized ran ecosystem,

    A. Garcia-Saavedra and X. Costa-Perez, “O-ran: Disrupting the virtual- ized ran ecosystem,” IEEE Communications Standards Magazine, vol. 5, no. 4, pp. 96–103, 2021

    2021

  10. [16]

    O-ran-enabled intelligent network slicing to meet service-level agreement (sla),

    J. Dai, L. Li, R. Safavinejad, S. Mahboob, H. Chen, V . V . Ratnam, H. Wang, J. Zhang, and L. Liu, “O-ran-enabled intelligent network slicing to meet service-level agreement (sla),” IEEE Transactions on Mobile Computing, 2024

    2024

  11. [17]

    Demystifying resource allocation policies in operational 5g mmwave networks,

    P. Dinh, M. Ghoshal, Y . Han, Y . Feng, D. Koutsonikolas, and J. Widmer, “Demystifying resource allocation policies in operational 5g mmwave networks,” IEEE Transactions on Networking , 2025

    2025

  12. [18]

    Multiverse at the edge: interacting real world and digital twins for wireless beamforming,

    B. Salehi, U. Demir, D. Roy, S. Pradhan, J. Dy, S. Ioannidis, and K. Chowdhury, “Multiverse at the edge: interacting real world and digital twins for wireless beamforming,” IEEE/ACM Transactions on Networking, 2024

    2024

  13. [19]

    Learning and reconstructing conflicts in o-ran: A graph neural network approach,

    A. Zolghadr, J. F. Santos, L. A. DaSilva, and J. Kibiłda, “Learning and reconstructing conflicts in o-ran: A graph neural network approach,” arXiv preprint arXiv:2412.14119 , 2024

    2024 arXiv

  14. [21]

    Joint energy-efficient and throughput- sufficient transmissions in 5g cells with deep q-learning,

    S. T. Spantideas, A. E. Giannopoulos, N. C. Kapsalis, A. Kalafatelis, C. N. Capsalis, and P. Trakadas, “Joint energy-efficient and throughput- sufficient transmissions in 5g cells with deep q-learning,” in 2021 IEEE International Mediterranean Conference on Communications and ...

    2021

  15. [23]

    Long-term throughput maxi- mization in wireless powered communication networks: A multi-task drl approach,

    A. Ahmadian, W. Shin, and H. Park, “Long-term throughput maxi- mization in wireless powered communication networks: A multi-task drl approach,” IEEE Internet of Things Journal , 2024

    2024

  16. [24]

    Sum rate maximization in iot networks with diversity-enhanced energy harvesting: A drl-guided approach,

    S. A. Ullah, M. A. Sohail, H. Jung, M. O. B. Saeed, and S. A. Hassan, “Sum rate maximization in iot networks with diversity-enhanced energy harvesting: A drl-guided approach,” IEEE Internet of Things Journal , 2024

    2024

  17. [25]

    Digital twin for o-ran toward 6g,

    H. X. Nguyen, K. Sun, D. To, Q.-T. Vien, and T. A. Le, “Digital twin for o-ran toward 6g,” IEEE Communications Magazine , 2024

    2024

  18. [26]

    3gpp standardized 5g channel model for iiot scenarios: A survey,

    T. Jiang, J. Zhang, P. Tang, L. Tian, Y . Zheng, J. Dou, H. Asplund, L. Raschkowski, R. D’Errico, and T. J ¨ams¨a, “3gpp standardized 5g channel model for iiot scenarios: A survey,” IEEE Internet of Things Journal, vol. 8, no. 11, pp. 8799–8815, 2021

    2021

  19. [27]

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

    3GPP TR 38.901 (Release 17), “Study on channel model for frequencies from 0.5 to 100 ghz,” 2022

    2022

  20. [28]

    Achieving energy efficiency in open radio access networks (oran) using xapps,

    S. S. Malik, S. Sil, K. Ranjan, M. S. Duhan, and A. Singh, “Achieving energy efficiency in open radio access networks (oran) using xapps,” Quantum Scientific & Academic Research Consortium , vol. 1, no. 1, 2024

    2024

  21. [29]

    Towards efficient conflict mitigation in the converged 6g open ran control plane,

    M. Corici, R. Modroiu, F. Eichhorn, E. Troudt, and T. Magedanz, “Towards efficient conflict mitigation in the converged 6g open ran control plane,” Annals of Telecommunications, pp. 1–11, 2024

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.