pith. sign in

arxiv: 2605.01112 · v1 · submitted 2026-05-01 · 💻 cs.NI

AIIM: Adaptive Inter-cell Interference Mitigation for Heterogeneous Multi-vendor 5G O-RAN Networks

Samuel Reinders , Alireza Ebrahimi Dorcheh , Ryan Barker , Tolunay Seyfi , Fatemeh Afghah This is my paper

Pith reviewed 2026-05-09 18:04 UTC · model grok-4.3

classification 💻 cs.NI
keywords O-RANinter-cell interference mitigationxAppPRB allocation5G networksQoSheterogeneous networksnear-RT RIC
0
0 comments X

The pith

An O-RAN xApp can coordinate PRB allocation across cells to improve QoS satisfaction and reduce interference losses while preserving throughput.

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

This paper develops and tests AIIM, a learning-driven xApp for the O-RAN near-real-time RIC that adapts physical resource block allocations among neighboring base stations. It accounts for overlapping resource use, user quality demands, and varying signal strengths in a mixed hardware and software testbed. A sympathetic reader would care because inter-cell interference limits performance in crowded 5G areas, and O-RAN's open interfaces enable such intelligent coordination without requiring full vendor alignment. If successful, the approach shows that data-driven control can deliver measurable gains in user satisfaction without sacrificing network capacity. The evaluation uses a hybrid setup to balance realism and practicality in multi-cell experiments.

Core claim

AIIM models the overlapping PRB regions across neighboring cells and learns coordinated allocation policies that adapt to per-user QoS demand and pathloss variation. When deployed in the hybrid experimental platform, it improves QoS satisfaction and reduces interference-induced PRB loss relative to proportional-fair scheduling baselines while maintaining comparable aggregate network throughput.

What carries the argument

The AIIM xApp, which learns coordinated physical resource block allocation policies across multiple base stations by modeling overlaps and adapting to traffic and channel conditions.

If this is right

  • More users meet their quality-of-service targets in dense multi-cell settings.
  • Physical resource block losses from interference decline without reducing total data throughput.
  • Interference management works across different vendor equipment through the O-RAN controller.
  • Hybrid test platforms enable reproducible studies of multi-cell coordination at lower cost.

Where Pith is reading between the lines

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

  • This method of adaptive allocation could extend to other resource management tasks in the RAN intelligent controller.
  • Operators might achieve higher effective capacity in existing infrastructure rather than adding more cells.
  • Further tests with diverse real-world traffic would confirm how well the policies transfer to new situations.

Load-bearing premise

The hybrid platform combining SDR-based and virtual base stations and users sufficiently captures the physical layer interactions and coordination constraints present in real heterogeneous multi-vendor 5G O-RAN networks.

What would settle it

A trial on a commercial multi-vendor 5G deployment where AIIM fails to increase the fraction of satisfied users or to lower PRB losses would show the claim does not hold outside the testbed.

Figures

Figures reproduced from arXiv: 2605.01112 by Alireza Ebrahimi Dorcheh, Fatemeh Afghah, Ryan Barker, Samuel Reinders, Tolunay Seyfi.

Figure 1
Figure 1. Figure 1: Virtual multi-vendor system model for per-cell PRB alloca view at source ↗
Figure 2
Figure 2. Figure 2: Frequency-domain PRB distribution for gNB2 (blue, virtual), view at source ↗
Figure 3
Figure 3. Figure 3: Testbed architecture with hardware connections and software view at source ↗
Figure 4
Figure 4. Figure 4: Performance metrics across PPO, DQN, SA-VA-PF, SA-CA-PF, and NSA-PF. view at source ↗
read the original abstract

Inter-cell interference is a persistent issue in dense 5G deployments, especially in heterogeneous Open Radio Access Network (O-RAN) environments where coordination between base stations is limited. This paper presents AIIM, an adaptive inter-cell interference mitigation xApp for the O-RAN near-real-time RAN Intelligent Controller (near-RT RIC) that performs coordinated physical resource block (PRB) allocation across multiple base stations under diverse traffic demands and channel conditions. Unlike prior studies that rely primarily on simulation or fully hardware-centric testbeds, AIIM is developed and evaluated in a full-stack O-RAN system built on srsRAN, Open5GS, and O-RAN Software Community (ORAN-SC), and deployed on a hybrid experimental platform that simultaneously combines software defined radio (SDR)-based and virtual gNodeBs (gNBs) and user equipment (UEs). This design preserves realistic PHY-layer interactions while substantially improving scalability, reproducibility, and cost-effectiveness for multi-cell interference experiments. AIIM explicitly models overlapping PRB regions across neighboring cells and learns coordinated allocation policies that adapt to per-user QoS demand and pathloss variation across the network. Experimental results show that AIIM improves QoS satisfaction and reduces interference-induced PRB loss relative to proportional-fair scheduling baselines while maintaining comparable aggregate network throughput. These results demonstrate the promise of scalable, learning-driven O-RAN control for practical interference management in heterogeneous multi-gNB 5G networks.\footnote{A video demonstration of the running system can be found at https://github.com/sireinders/AIIM-Multi-gNB-Interference.git.}

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript presents AIIM, an xApp for the O-RAN near-RT RIC that performs adaptive, coordinated physical resource block (PRB) allocation to mitigate inter-cell interference across multiple gNBs in heterogeneous 5G networks. It is developed and evaluated on a hybrid experimental platform combining SDR-based and virtual gNBs/UEs using the srsRAN, Open5GS, and ORAN-SC stack. The central experimental claim is that the learned policies improve QoS satisfaction and reduce interference-induced PRB loss relative to proportional-fair baselines while maintaining comparable aggregate throughput.

Significance. If the results hold under more rigorous validation, the work is significant as a practical demonstration of near-RT RIC xApps for interference management in O-RAN. The hybrid SDR-virtual testbed approach is a notable strength for balancing PHY realism with scalability and reproducibility, offering a template for future multi-cell O-RAN experiments that avoids the cost of fully hardware-centric setups. It contributes concrete evidence that learning-driven coordination can deliver measurable QoS gains without throughput penalties.

major comments (3)
  1. [Abstract] Abstract: The headline performance claims (improved QoS satisfaction, reduced PRB loss, comparable throughput) are stated without any quantitative metrics, effect sizes, error bars, number of experimental runs, or statistical tests. This absence makes it impossible to assess the practical magnitude or reliability of the gains and is load-bearing for the central experimental result.
  2. [Experimental setup] Hybrid experimental platform description: The evaluation rests on the assumption that the hybrid SDR-virtual platform (srsRAN/ORAN-SC with idealized virtual components) sufficiently captures PHY-layer interactions, HARQ timing, and coordination latencies of real heterogeneous multi-vendor deployments. No analysis or ablation is provided to show that omitted vendor-specific impairments do not alter the learned policy's effectiveness, undermining transferability of the reported improvements.
  3. [Title and Abstract] Title and abstract: The work is framed as addressing 'heterogeneous multi-vendor' O-RAN networks, yet the platform is constructed entirely from open-source components (srsRAN, ORAN-SC) rather than proprietary vendor stacks. This mismatch is load-bearing for the generalizability claim, as inter-vendor E2 delays and scheduler differences are precisely the effects the skeptic note identifies as potentially dominant.
minor comments (2)
  1. [Method] The learning algorithm details (state representation, reward function, training procedure) are referenced but not fully specified in the provided text; adding pseudocode or a dedicated subsection would improve reproducibility.
  2. [Footnote] The GitHub footnote for the video demonstration should include a direct link to the exact commit or release used for the reported experiments to support reproducibility.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for the constructive feedback, which highlights opportunities to strengthen the quantitative presentation of results and clarify the scope and limitations of our experimental platform. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The headline performance claims (improved QoS satisfaction, reduced PRB loss, comparable throughput) are stated without any quantitative metrics, effect sizes, error bars, number of experimental runs, or statistical tests. This absence makes it impossible to assess the practical magnitude or reliability of the gains and is load-bearing for the central experimental result.

    Authors: We agree that the abstract should provide quantitative support for the central claims. The revised abstract will include specific effect sizes (e.g., relative improvements in QoS satisfaction and PRB loss), the number of experimental runs, and reference to statistical tests performed, drawn directly from the evaluation results in the manuscript body. revision: yes

  2. Referee: [Experimental setup] Hybrid experimental platform description: The evaluation rests on the assumption that the hybrid SDR-virtual platform (srsRAN/ORAN-SC with idealized virtual components) sufficiently captures PHY-layer interactions, HARQ timing, and coordination latencies of real heterogeneous multi-vendor deployments. No analysis or ablation is provided to show that omitted vendor-specific impairments do not alter the learned policy's effectiveness, undermining transferability of the reported improvements.

    Authors: The hybrid platform prioritizes PHY realism via SDR components while leveraging virtual elements for multi-cell scalability. We will add a dedicated discussion of the idealized virtual-component assumptions and their potential effects on HARQ timing and coordination. We will also expand the limitations section to explicitly note that vendor-specific impairments from proprietary stacks are not modeled. revision: partial

  3. Referee: [Title and Abstract] Title and abstract: The work is framed as addressing 'heterogeneous multi-vendor' O-RAN networks, yet the platform is constructed entirely from open-source components (srsRAN, ORAN-SC) rather than proprietary vendor stacks. This mismatch is load-bearing for the generalizability claim, as inter-vendor E2 delays and scheduler differences are precisely the effects the skeptic note identifies as potentially dominant.

    Authors: We acknowledge that the testbed relies on open-source components rather than proprietary multi-vendor stacks. We will revise the title to 'AIIM: Adaptive Inter-cell Interference Mitigation for Heterogeneous Multi-gNB 5G O-RAN Networks' and update the abstract to describe the platform as combining SDR-based and virtual gNBs/UEs to emulate heterogeneity within the O-RAN architecture, while tempering generalizability statements. revision: yes

standing simulated objections not resolved
  • Empirical demonstration that omitted vendor-specific impairments do not alter the learned policy effectiveness (this would require access to proprietary vendor stacks for additional ablation experiments)

Circularity Check

0 steps flagged

No circularity: claims rest on direct experimental comparison

full rationale

The paper describes development and evaluation of the AIIM xApp for coordinated PRB allocation in a hybrid O-RAN testbed. It reports measured improvements in QoS satisfaction and PRB loss versus proportional-fair baselines, with no equations, fitted parameters renamed as predictions, self-definitional loops, or load-bearing self-citations. The chain is system implementation followed by empirical validation on described hardware/software, which is self-contained against external benchmarks and does not reduce to its inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach assumes standard O-RAN interfaces allow xApps to control PRB allocation and that overlapping PRB regions can be explicitly modeled; no free parameters or new entities are described in the abstract.

axioms (1)
  • domain assumption O-RAN near-RT RIC xApps can perform coordinated PRB allocation across multiple gNBs
    Invoked as the foundation for the AIIM xApp design in the abstract.

pith-pipeline@v0.9.0 · 5617 in / 1316 out tokens · 42314 ms · 2026-05-09T18:04:54.436338+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages

  1. [1]

    An overview on intercell interference manage- ment in mobile cellular networks: From 2G to 5G,

    Y. Zhouet al., “An overview on intercell interference manage- ment in mobile cellular networks: From 2G to 5G,” in2014 IEEE International Conference on Communication Systems, 2014, pp. 217–221

    work page 2014

  2. [2]

    Ra- diotwin: A digital building material twin for wideband, cross-link, cross-band wireless channel prediction,

    M. I. Rochmanet al., “Neutral-hosts in the shared mid-bands: Addressing indoor cellular performance,” in2025 IEEE Inter- national Symposium on Dynamic Spectrum Access Networks (DySPAN). IEEE, May 2025, p. 1–9. [Online]. Available: http://dx.doi.org/10.1109/DySPAN64764.2025.11115898

    work page doi:10.1109/dyspan64764.2025.11115898 2025

  3. [3]

    A comprehensive analysis of secondary co- existence in a real-world cbrs deployment,

    A. Tushaet al., “A comprehensive analysis of secondary co- existence in a real-world cbrs deployment,” in2024 IEEE In- ternational Symposium on Dynamic Spectrum Access Networks (DySPAN), 2024, pp. 79–87

    work page 2024

  4. [4]

    Understanding O-RAN: Architecture, in- terfaces, algorithms, security, and research challenges,

    M. Poleseet al., “Understanding O-RAN: Architecture, in- terfaces, algorithms, security, and research challenges,”IEEE Communications Surveys & Tutorials, vol. 25, no. 2, pp. 1376– 1411, 2023

    work page 2023

  5. [5]

    Toward next generation open radio access networks: What O-RAN can and cannot do!

    A. S. Abdallaet al., “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

  6. [6]

    REAL: Reinforcement learning-enabled xApps for experimental closed-loop optimization in O-RAN with OSC RIC and srsRAN,

    R. Barkeret al., “REAL: Reinforcement learning-enabled xApps for experimental closed-loop optimization in O-RAN with OSC RIC and srsRAN,” in2025 IEEE International Conference on Communications Workshops (ICC Workshops). IEEE, Jun. 2025, p. 389–395. [Online]. Available: http: //dx.doi.org/10.1109/ICCWorkshops67674.2025.11162144

    work page doi:10.1109/iccworkshops67674.2025.11162144 2025

  7. [7]

    ChARM: NextG spectrum sharing through data-driven real-time O-RAN dy- namic control,

    L. Baldesi, F. Restuccia, and T. Melodia, “ChARM: NextG spectrum sharing through data-driven real-time O-RAN dy- namic control,” inIEEE INFOCOM 2022 - IEEE Conference on Computer Communications, 2022, pp. 240–249

    work page 2022

  8. [8]

    Prompt-tuned LLM- augmented DRL for dynamic O-RAN network slicing,

    F. Lotfi, H. Rajoli, and F. Afghah, “Prompt-tuned LLM- augmented DRL for dynamic O-RAN network slicing,” 2025. [Online]. Available: https://arxiv.org/abs/2506.00574

    work page arXiv 2025

  9. [9]

    Interference management issues for the future 5G network: A review - telecommunication systems,

    F. Qamaret al., “Interference management issues for the future 5G network: A review - telecommunication systems,” May

    work page

  10. [10]

    Available: https://link.springer.com/article/10

    [Online]. Available: https://link.springer.com/article/10. 1007/s11235-019-00578-4

    work page

  11. [11]

    Algorithms for enhanced inter-cell interference coordination (eICIC) in LTE HetNets,

    S. Debet al., “Algorithms for enhanced inter-cell interference coordination (eICIC) in LTE HetNets,”IEEE/ACM Transac- tions on Networking, vol. 22, no. 1, pp. 137–150, 2014

    work page 2014

  12. [12]

    Coordinated multipoint transmission and recep- tion in LTE-advanced: deployment scenarios and operational challenges,

    D. Leeet al., “Coordinated multipoint transmission and recep- tion in LTE-advanced: deployment scenarios and operational challenges,”IEEE Communications Magazine, vol. 50, no. 2, pp. 148–155, 2012

    work page 2012

  13. [13]

    Evolution of limited-feedback CoMP systems from 4G to 5G: CoMP features and limited-feedback ap- proaches,

    Q. Cuiet al., “Evolution of limited-feedback CoMP systems from 4G to 5G: CoMP features and limited-feedback ap- proaches,”IEEE Vehicular Technology Magazine, vol. 9, no. 3, pp. 94–103, 2014

    work page 2014

  14. [14]

    DORA: Dynamic O- RAN resource allocation for multi-slice 5G networks,

    A. E. Dorcheh, T. Seyfi, and F. Afghah, “DORA: Dynamic O-RAN resource allocation for multi-slice 5G networks,” 2025. [Online]. Available: https://arxiv.org/abs/2509.07242

    work page arXiv 2025

  15. [15]

    ORANSlice: An open-source 5G network slicing platform for O-RAN,

    H. Chenget al., “ORANSlice: An open-source 5G network slicing platform for O-RAN,” 2024. [Online]. Available: https: //arxiv.org/abs/2410.12978

    work page arXiv 2024

  16. [16]

    xDiff: Online diffusion model for collaborative inter-cell interference management in 5G O- RAN,

    P. Yan, H. Zeng, and Y. T. Hou, “xDiff: Online diffusion model for collaborative inter-cell interference management in 5G O- RAN,”IEEE Transactions on Networking, vol. 34, pp. 1363– 1376, 2026

    work page 2026

  17. [17]

    Interference-aware intelligent scheduling for virtualized private 5G networks,

    B. Akgunet al., “Interference-aware intelligent scheduling for virtualized private 5G networks,”IEEE Access, vol. PP, pp. 1– 1, 01 2024

    work page 2024

  18. [18]

    A machine learning-based approach for interference mitigation to enhance QoS and QoE in 5G O-RAN networks,

    D. Anand, M. A. Togou, and G.-M. Muntean, “A machine learning-based approach for interference mitigation to enhance QoS and QoE in 5G O-RAN networks,” in2024 IEEE 35th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 2024, pp. 1–6

    work page 2024

  19. [19]

    Colosseum: Large-scale wireless experimenta- tion through hardware-in-the-loop network emulation,

    L. Bonatiet al., “Colosseum: Large-scale wireless experimenta- tion through hardware-in-the-loop network emulation,” in2021 IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN), 2021, pp. 113–122

    work page 2021

  20. [20]

    AERPAW emulation overview and prelimi- nary performance evaluation,

    A. Panickeret al., “AERPAW emulation overview and prelimi- nary performance evaluation,”Elsevier Computer Communica- tions, vol. 194, pp. 158–169, 2021

    work page 2021

  21. [21]

    Open AI Cellular (OAIC): An open source 5G O-RAN testbed for design and testing of AI-based RAN management algorithms,

    H. Zhanget al., “Open AI Cellular (OAIC): An open source 5G O-RAN testbed for design and testing of AI-based RAN management algorithms,”IEEE Network, vol. 37, no. 6, pp. 102–109, 2023

    work page 2023

  22. [22]

    Study on Sce- narios and Requirements for Next Generation Access Technolo- gies,

    3rd Generation Partnership Project (3GPP), “Study on Sce- narios and Requirements for Next Generation Access Technolo- gies,” 3rd Generation Partnership Project (3GPP), Technical Report (TR) 38.913, 2020, release 16. [Online]. Available: https://www.3gpp.org/ftp/Specs/archive/38 series/38.913/

    work page 2020