pith. sign in

arxiv: 2605.15446 · v1 · pith:ONLAZOZWnew · submitted 2026-05-14 · 💻 cs.NI

Resilience under Uncertainty: Securing 6G through Stochastic Reinstantiation of RAN Functions

Gabriel Almeida , Jacek Kibi{\l}da , Joao F. Santos , Kleber Vieira Cardoso This is my paper

Pith reviewed 2026-05-19 14:28 UTC · model grok-4.3

classification 💻 cs.NI
keywords 6GRAN disaggregationresiliencestochastic optimizationcascading failuresnetwork recoverycloud reinstantiation
0
0 comments X

The pith

Disaggregated 6G networks recover from cascading RAN failures by reinstantiating CUs and DUs in alternative cloud locations using two-stage stochastic optimization under uncertainty.

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

The paper proposes a resilience mechanism that reacts to failures in Radio Units by reinstantiating Central Units and Distributed Units in alternative cloud locations to restore function chains. It models the recovery as a two-stage stochastic optimization problem in which reinstantiation and routing decisions are made before uncertainty is resolved and bandwidth allocation occurs afterward. The problem is solved with Sample Average Approximation to create a tractable deterministic equivalent. Numerical tests on real-world topologies across failure scenarios and traffic conditions show up to 80 percent higher recovery performance than conventional mechanisms. A sympathetic reader cares because disaggregation of base stations creates new cascading failure risks that standard resilience approaches handle poorly.

Core claim

We propose the first resilience mechanism for disaggregated mobile networks that leverages the adaptive reinstantiation of RAN functions under uncertainty to mitigate disruptions and maintain service continuity in the presence of compromised infrastructure. Our mechanism reacts to cascading failures that disrupt Radio Units by reinstantiating Central Units and Distributed Units in alternative cloud locations, restoring their function chains while accounting for uncertainty in users' locations and wireless channel conditions during the in-failure state. We formulate this recovery process as a two-stage stochastic optimization problem, where reinstantiation and routing decisions are made under

What carries the argument

Two-stage stochastic optimization problem solved via Sample Average Approximation, with reinstantiation and routing decisions under uncertainty in the first stage and bandwidth allocation after uncertainty is resolved in the second stage.

If this is right

  • The approach restores disrupted function chains for RUs by selecting alternative locations for CUs and DUs.
  • It explicitly models uncertainty in user locations and channel conditions through a finite scenario set.
  • It delivers up to 80 percent higher recovery performance than conventional resilience mechanisms in the evaluated cases.
  • The same formulation applies across varied failure scenarios and traffic demand conditions on real-world topologies.

Where Pith is reading between the lines

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

  • Similar stochastic reinstantiation could be tested for resilience in other disaggregated systems such as edge computing clusters.
  • Replacing the fixed scenario set with online learning to generate scenarios dynamically might improve adaptation to changing conditions.
  • Deploying the mechanism at scale would require verifying that sufficient alternative cloud capacity exists in practice.

Load-bearing premise

Alternative cloud locations for reinstantiating CUs and DUs are always available and uncertainty in users' locations and wireless channel conditions during failure can be adequately captured by a finite set of scenarios in the stochastic model.

What would settle it

Running recovery tests on a live disaggregated network testbed with real cascading RU failures, restricted alternative locations, and measured channel conditions to check whether the performance gain over conventional mechanisms still reaches 80 percent.

Figures

Figures reproduced from arXiv: 2605.15446 by Gabriel Almeida, Jacek Kibi{\l}da, Joao F. Santos, Kleber Vieira Cardoso.

Figure 1
Figure 1. Figure 1: Disaggregated mobile networks are susceptible to [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Illustrative communication resilience dynamics show [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Impact of failures on users’ connectivity. Non-affected [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Representation of our two-stage stochastic optimization [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Impact of the failure scenarios with different severity [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 9
Figure 9. Figure 9: Comparison of the distributions of throughput resilience [PITH_FULL_IMAGE:figures/full_fig_p010_9.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison of the temporal evolution of network utility [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 11
Figure 11. Figure 11: Comparison of recovery performance across different [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
Figure 10
Figure 10. Figure 10: Recovery performance of different resilience mecha [PITH_FULL_IMAGE:figures/full_fig_p011_10.png] view at source ↗
read the original abstract

The disaggregation of base stations into discrete RAN functions introduces new threats to mobile networks, as failures in one RAN function can trigger cascading failures and interrupt entire function chains, with potential to degrade network performance and disrupt service. In this paper, we propose the first resilience mechanism for disaggregated mobile networks that leverages the adaptive reinstantiation of RAN functions under uncertainty to mitigate disruptions and maintain service continuity in the presence of compromised infrastructure. Our mechanism reacts to cascading failures that disrupt Radio Units (RUs) by reinstantiating Central Units (CUs) and Distributed Units (DUs) in alternative cloud locations, restoring their function chains while accounting for uncertainty in users' locations and wireless channel conditions during the in-failure state. We formulate this recovery process as a two-stage stochastic optimization problem, where reinstantiation and routing decisions are made under uncertainty, and bandwidth allocation decisions are performed after uncertainty is resolved. We solve the problem using a Sample Average Approximation (SAA)-based solution as a tractable, deterministic equivalent problem. We numerically evaluate our approach on a real-world disaggregated mobile network topology across multiple failure scenarios and traffic demand conditions, and our results demonstrate that our solution can achieve up to 80% higher recovery performance compared to conventional resilience mechanisms.

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

2 major / 2 minor

Summary. The paper proposes the first resilience mechanism for disaggregated 6G RANs that uses adaptive reinstantiation of CUs and DUs in alternative cloud locations to recover from cascading RU failures. It formulates the recovery as a two-stage stochastic optimization problem (first-stage reinstantiation and routing under uncertainty in user locations and channels; second-stage bandwidth allocation after uncertainty resolves), solves it via Sample Average Approximation (SAA), and evaluates the approach on a real-world topology to claim up to 80% higher recovery performance versus conventional mechanisms.

Significance. If the modeling assumptions hold, the work contributes a novel application of two-stage stochastic programming to RAN function-chain resilience, with the SAA solution providing a tractable deterministic equivalent and the real-topology evaluation offering concrete performance numbers. This could inform practical 6G deployment strategies for maintaining service continuity under infrastructure compromise.

major comments (2)
  1. [Abstract / two-stage stochastic formulation] Abstract and formulation section: The two-stage stochastic model and the reported 80% recovery gain rest on the assumption that alternative cloud locations for reinstantiating CUs and DUs are always available and non-saturated. The manuscript provides no analysis or fallback for the case in which the candidate set is empty or fully occupied, which directly affects the validity of the first-stage decisions and the performance comparison to conventional mechanisms.
  2. [Numerical evaluation] Numerical evaluation section: The SAA solution relies on a finite set of scenarios to approximate uncertainty in user locations and wireless channel conditions during failure. No sensitivity analysis to the number of scenarios, no verification that the sample captures tail events, and no error bars or confidence intervals are reported, limiting the ability to confirm that the 80% improvement is robust rather than an artifact of the chosen scenario set.
minor comments (2)
  1. [Numerical evaluation] The description of the conventional resilience baselines used for comparison is not detailed enough to allow exact reproduction of the 80% figure.
  2. [System model / evaluation setup] The manuscript does not discuss how the real-world topology was mapped to the disaggregated RAN model or how traffic demand conditions were generated.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major comment below, indicating the specific revisions we will incorporate to improve the manuscript.

read point-by-point responses

  1. Referee: [Abstract / two-stage stochastic formulation] Abstract and formulation section: The two-stage stochastic model and the reported 80% recovery gain rest on the assumption that alternative cloud locations for reinstantiating CUs and DUs are always available and non-saturated. The manuscript provides no analysis or fallback for the case in which the candidate set is empty or fully occupied, which directly affects the validity of the first-stage decisions and the performance comparison to conventional mechanisms.

    Authors: We acknowledge that the current two-stage formulation implicitly assumes a non-empty and non-saturated set of candidate cloud locations for CU/DU reinstantiation. This modeling choice reflects the multi-cloud and edge-computing infrastructure expected in 6G disaggregated RANs. To directly address the concern, we will revise the formulation section to explicitly state the assumption, discuss its implications for first-stage decisions, and introduce a fallback mechanism (e.g., partial recovery or deferral to baseline mechanisms) when the candidate set is empty or saturated. We will also add a sensitivity study in the numerical evaluation that varies the number and occupancy of candidate locations and reports the resulting impact on recovery performance. revision: yes

  2. Referee: [Numerical evaluation] Numerical evaluation section: The SAA solution relies on a finite set of scenarios to approximate uncertainty in user locations and wireless channel conditions during failure. No sensitivity analysis to the number of scenarios, no verification that the sample captures tail events, and no error bars or confidence intervals are reported, limiting the ability to confirm that the 80% improvement is robust rather than an artifact of the chosen scenario set.

    Authors: We agree that additional validation of the SAA approximation is warranted. In the revised manuscript we will include (i) a sensitivity analysis over scenario counts (e.g., 50, 100, 200), (ii) explicit verification that the scenario-generation procedure includes tail events for user locations and channel conditions, and (iii) error bars together with 95% confidence intervals on all reported performance metrics. These additions will demonstrate that the observed recovery gains, including the up to 80% improvement, remain stable across different sample sizes and are not artifacts of the particular scenario set. revision: yes

Circularity Check

0 steps flagged

No significant circularity: new two-stage stochastic formulation evaluated on external topology

full rationale

The paper introduces a two-stage stochastic optimization model for RAN function reinstantiation under uncertainty in user locations and channel conditions, solved via Sample Average Approximation as a deterministic equivalent. Recovery performance is assessed through numerical experiments on a real-world disaggregated mobile network topology across multiple failure scenarios and traffic conditions, with comparisons to conventional mechanisms. No load-bearing step reduces by construction to fitted inputs, self-definitions, or self-citation chains; the central claims rest on the external validity of the modeled scenarios and topology data rather than internal reparameterization.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard stochastic programming techniques applied to a networking domain; no new physical entities are introduced, but the model depends on scenario-based uncertainty representation and availability of alternative cloud resources.

free parameters (1)
  • Number of scenarios for SAA approximation
    Used to create the deterministic equivalent; specific count not stated in abstract but required for tractability.
axioms (1)
  • domain assumption Uncertainty in user locations and wireless channel conditions during failure states can be represented by a discrete set of scenarios
    Invoked to enable the two-stage stochastic formulation where first-stage decisions precede uncertainty resolution.

pith-pipeline@v0.9.0 · 5767 in / 1325 out tokens · 53162 ms · 2026-05-19T14:28:09.918265+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages

  1. [1]

    Managing O-RAN Networks: xApp Develop- ment from Zero to Hero,

    J. F. Santoset al., “Managing O-RAN Networks: xApp Develop- ment from Zero to Hero,”IEEE Communications Surveys & Tutorials (COMST), vol. 28, pp. 800–840, 2026

    work page 2026

  2. [2]

    PlaceRAN: Optimal Placement of Virtualized Network Functions in Beyond 5G Radio Access Networks,

    F. Z. Moraiset al., “PlaceRAN: Optimal Placement of Virtualized Network Functions in Beyond 5G Radio Access Networks,”IEEE Transactions on Mobile Computing (TMC), vol. 22, no. 9, pp. 5434– 5448, 2023

    work page 2023

  3. [3]

    Sustainability and Power Outage-aware Place- ment of Edge Computing in NextG RANs,

    K. Sai and D. Tipper, “Sustainability and Power Outage-aware Place- ment of Edge Computing in NextG RANs,” inIEEE International Symposium on Technology and Society (ISTAS), 2024, pp. 1–8

    work page 2024

  4. [4]

    Learning and Reconstructing Conflicts in O-RAN: A Graph Neural Network Approach,

    A. Zolghadret al., “Learning and Reconstructing Conflicts in O-RAN: A Graph Neural Network Approach,” inIEEE Wireless Communications and Networking Conference (WCNC), 2025, pp. 01–06

    work page 2025

  5. [5]

    Resilient Baseband Processing in Virtualized RANs with Slingshot,

    N. Lazarevet al., “Resilient Baseband Processing in Virtualized RANs with Slingshot,” inACM Special Interest Group on Data Communica- tions (SIGCOMM), 2023, pp. 654–667

    work page 2023

  6. [6]

    Adaptive Reallocation of RAN Functions for Resilient 6G Networks,

    G. M. Almeidaet al., “Adaptive Reallocation of RAN Functions for Resilient 6G Networks,” 2025. [Online]. Available: https: //arxiv.org/abs/2511.08467

    work page arXiv 2025

  7. [7]

    Resilience and criticality: Brothers in arms for 6G,

    R.-J. Reifertet al., “Resilience and Criticality: Brothers in Arms for 6G,” 2024. [Online]. Available: https://arxiv.org/abs/2412.03661

    work page arXiv 2024

  8. [8]

    Optimizing Energy Consumption for vRAN Placement in O-RAN Systems With Flexible Transport Networks,

    W. T. Pires-JRet al., “Optimizing Energy Consumption for vRAN Placement in O-RAN Systems With Flexible Transport Networks,”IEEE Open Journal of the Communications Society (OJ-COMS), vol. 6, pp. 4279–94, 2025

    work page 2025

  9. [9]

    Cost-Aware Neural Adaptive Scaling for vRAN Resource Allocation,

    T. Kimet al., “Cost-Aware Neural Adaptive Scaling for vRAN Resource Allocation,”IEEE Transactions on Mobile Computing (TMC), pp. 1–11, 2025

    work page 2025

  10. [10]

    GreenRAN: A Channel-Aware Green O-RAN Frame- work for NextG Mobile Systems,

    C. Youet al., “GreenRAN: A Channel-Aware Green O-RAN Frame- work for NextG Mobile Systems,” inIEEE Conference on Computer Communications (INFOCOM), 2025

    work page 2025

  11. [11]

    Multi-Resource Orchestration and Energy-Aware VNF Placement for Open RAN,

    H. Hojeijet al., “Multi-Resource Orchestration and Energy-Aware VNF Placement for Open RAN,” inIEEE International Conference on Network Softwarization (NetSoft), 2025

    work page 2025

  12. [12]

    Optimal Resource Allocation With Delay Guarantees for Network Slicing in Disaggregated RAN,

    F. G. C. Rochaet al., “Optimal Resource Allocation With Delay Guarantees for Network Slicing in Disaggregated RAN,”IEEE/ACM Transactions on Networking (TNET), vol. 34, pp. 4249–4268, 2026

    work page 2026

  13. [13]

    Resilient Deployment of Virtual Network Functions,

    M. Sch ¨olleret al., “Resilient Deployment of Virtual Network Functions,” inInternational Congress on Ultra Modern Telecom. and Control Systems (ICUMT), 2013, pp. 208–214

    work page 2013

  14. [14]

    Minimum-Cost Virtual Network Function Resilience,

    Y . Carlinetet al., “Minimum-Cost Virtual Network Function Resilience,” inINOC 2019, Avignon, France, Jun. 2019

    work page 2019

  15. [15]

    Multi-Agent Deep Reinforcement Learning for Resilience Optimization in 5G RAN,

    S. Kaadaet al., “Multi-Agent Deep Reinforcement Learning for Resilience Optimization in 5G RAN,” 2024. [Online]. Available: https://arxiv.org/abs/2407.18066

    work page arXiv 2024

  16. [16]

    Comeback Kid: Resilience for Mixed-Critical Wire- less Network Resource Management,

    R.-J. Reifertet al., “Comeback Kid: Resilience for Mixed-Critical Wire- less Network Resource Management,”IEEE Transactions on Vehicular Technology (TVT), vol. 72, no. 12, pp. 16 177–16 194, 2023

    work page 2023

  17. [17]

    RIC-O: Efficient Placement of a Disaggregated and Distributed RAN Intelligent Controller With Dynamic Clustering of Radio Nodes,

    G. M. Almeidaet al., “RIC-O: Efficient Placement of a Disaggregated and Distributed RAN Intelligent Controller With Dynamic Clustering of Radio Nodes,”IEEE Journal on Selected Areas in Communications (JSAC), vol. 42, no. 2, pp. 446–459, 2024

    work page 2024

  18. [18]

    The R3 Concept: Reliability, Robustness, and Resilience,

    G. Zissis, “The R3 Concept: Reliability, Robustness, and Resilience,” IEEE Industry Applications Magazine (IAS), vol. 25, no. 4, pp. 5–6, 2019

    work page 2019

  19. [19]

    (2022, Aug.) NextG Alliance Report: Trust, Security, and Resilience for 6G Systems

    ATIS. (2022, Aug.) NextG Alliance Report: Trust, Security, and Resilience for 6G Systems. NextG Alliance. Accessed on Oct. 01,

    work page 2022

  20. [20]

    Available: https://www.nextgalliance.org/white\ papers/ trust-security-and-resilience-for-6g-systems

    [Online]. Available: https://www.nextgalliance.org/white\ papers/ trust-security-and-resilience-for-6g-systems

    work page

  21. [21]

    (2023, Oct.) Key Strategies for 6G Smart Networks and Services

    6G IA. (2023, Oct.) Key Strategies for 6G Smart Networks and Services. 6G IA. Accessed on Oct. 01, 2023. [Online]. Available: https://6g-ia. eu/wp-content/uploads/2023/10/6g-ia-position-paper\ 2023\ final.pdf

    work page 2023

  22. [22]

    5GRecon: Automating 5G network function recogni- tion and misconfiguration troubleshooting,

    S. Da Canalet al., “5GRecon: Automating 5G network function recogni- tion and misconfiguration troubleshooting,” inIFIP/IEEE International Conference on Performance Evaluation and Modeling in Wired and Wireless Networks (PEMWN), 2024, pp. 1–6

    work page 2024

  23. [23]

    Failure analysis in next-generation critical cellular com- munication infrastructures,

    S. Biet al., “Failure analysis in next-generation critical cellular com- munication infrastructures,”arXiv preprint arXiv:2402.04448, 2024

    work page arXiv 2024

  24. [24]

    Is my Internet Down?: Sifting through User-affecting Outages with Google Trends,

    E. C. Kirciet al., “Is my Internet Down?: Sifting through User-affecting Outages with Google Trends,” inACM Internet Measurement Conference (IMC, ser. IMC ’22, New York, NY , USA, 2022, p. 290–297

    work page 2022

  25. [25]

    Enabling Resilience in Virtualized RANs with Atlas,

    J. Xinget al., “Enabling Resilience in Virtualized RANs with Atlas,” inACM Annual International Conference on Mobile Computing and Networking (MobiCom), 2023, pp. 1–15

    work page 2023

  26. [26]

    On Optimal Orchestration of Virtualized Cellular Networks With Statistical Multiplexing,

    S. Chatterjeeet al., “On Optimal Orchestration of Virtualized Cellular Networks With Statistical Multiplexing,”IEEE Transactions on Wireless Communications (TWC), vol. 21, no. 1, pp. 310–325, 2022

    work page 2022

  27. [27]

    J. R. Birge and F. Louveaux,Introduction to Stochastic Programming. Springer, 1997

    work page 1997

  28. [28]

    Bi-Parameterized Two-Stage Stochastic Min-Max and Min-Min Mixed Integer Programs,

    S. Kang and M. Bansal, “Bi-Parameterized Two-Stage Stochastic Min-Max and Min-Min Mixed Integer Programs,” 2025. [Online]. Available: https://arxiv.org/abs/2501.01081

    work page arXiv 2025

  29. [29]

    Resilience analysis and quantification method for 5G-Radio Access Networks,

    Kaada, Soumeya and Alberi Morel, Marie Line and Rubino, Gerardo and Jelassi, Sofiene, “Resilience analysis and quantification method for 5G-Radio Access Networks,” inInternational Conference on Network of the Future (NoF), 2022, pp. 1–9

    work page 2022

  30. [30]

    Energy-Efficient Orchestration of Metro-Scale 5G Radio Access Networks,

    R. Singhet al., “Energy-Efficient Orchestration of Metro-Scale 5G Radio Access Networks,” inIEEE Conference on Computer Commu- nications (INFOCOM), 2021, pp. 1–10

    work page 2021

  31. [31]

    5G-crosshaul, D1.2: final 5G-crosshaul system design and economic analysis,

    5G-crosshaul Project, “5G-crosshaul, D1.2: final 5G-crosshaul system design and economic analysis,” 2017

    work page 2017

  32. [32]

    Propagation Path Loss Models for 5G Urban Micro- and Macro-Cellular Scenarios,

    S. Sunet al., “Propagation Path Loss Models for 5G Urban Micro- and Macro-Cellular Scenarios,” inIEEE Vehicular Technology Conference (VTC Spring), 2016, pp. 1–6

    work page 2016

  33. [33]

    Study on 3GPP Rural Macrocell Path Loss Models for Millimeter Wave Wireless Communications,

    G. R. MacCartney and T. S. Rappaport, “Study on 3GPP Rural Macrocell Path Loss Models for Millimeter Wave Wireless Communications,” in IEEE International Conference on Communications (ICC), 2017, pp. 1–7

    work page 2017

  34. [34]

    3rd Generation Partnership Project; 5G; NG-RAN; Architecture description,

    3GPP, “3rd Generation Partnership Project; 5G; NG-RAN; Architecture description,” 3GPP, Tech. Rep. TS 38.401 V19.1.0, 2025

    work page 2025

  35. [35]

    Cloud Architecture and Deployment Scenarios for O-RAN Virtualized RAN,

    Open RAN Alliance, “Cloud Architecture and Deployment Scenarios for O-RAN Virtualized RAN,” O-RAN Alliance, Tech. Rep., 2022

    work page 2022

  36. [36]

    Environmental Engineering (EE); Assessment of mobile network energy efficiency,

    ETSI, “Environmental Engineering (EE); Assessment of mobile network energy efficiency,” European Telecommunications Standards Institute (ETSI), Tech. Rep. ES 203 228 (V1.4.1), April 2022

    work page 2022

  37. [37]

    3rd Generation Partnership Project; Physical channels and mod- ulations,

    3GPP, “3rd Generation Partnership Project; Physical channels and mod- ulations,” 3GPP, Tech. Rep. TS 38.211 V19.0.0, 2025

    work page 2025

  38. [38]

    Understanding Mobile Traffic Patterns of Large Scale Cellular Towers in Urban Environment,

    F. Xuet al., “Understanding Mobile Traffic Patterns of Large Scale Cellular Towers in Urban Environment,”IEEE/ACM Transactions on Networking (TNET), vol. 25, no. 2, pp. 1147–1161, 2017

    work page 2017