Towards Smart Fronthauling Management: Experimental Insights from a 5G Testbed

Antonio Capone; Danilo De Donno; Eugenio Moro; Ilario Filippini; Marcello Morini; Salvatore Moscato

REVIEW 4 major objections 4 minor 16 references

Towards Smart Fronthauling Management: Experimental Insights from a 5G Testbed

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A live commercial 5G testbed confirms that a simple cell reconfiguration—cutting resource blocks and MIMO layers—can keep a wireless fronthaul link alive when capacity drops, and that the uplink, not the downlink, sets the real bottleneck.

desk verdict Valuable measurement data from a commercial 5G testbed, but the paper overclaims dynamic 'smart fronthauling' based on static experiments; send to review with a required reframing and data release. read the letter →

arxiv 2411.13989 v1 pith:HWGME5WV submitted 2024-11-21 cs.NI

Marcello Morini , Eugenio Moro , Ilario Filippini , Danilo De Donno , Salvatore Moscato , Antonio Capone This is my paper · ORCID

classification cs.NI

keywords fronthaulC-RANfunctionalspliteCPRImillimeterwave5GtestbedcellreconfigurationE-band

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

Centralized 5G networks depend on a fronthaul link whose capacity must always cover the radio access load; when that link is wireless, rain or congestion can temporarily cut capacity. This paper tries to establish that a simple management lever—reconfiguring the cell to fewer resource blocks and MIMO layers—can keep the connection alive, and that the fronthaul rate needed can be predicted by two linear formulas in terms of resource blocks, MIMO layers, modulation bits, and OFDM symbol duration. Using a commercial 5G millimeter-wave testbed with two E-band transceivers inserted into the fronthaul to emulate capacity reductions, the authors measure that halving the configured bandwidth roughly halves both the access throughput and the required fronthaul capacity, matching the model. They also find that the uplink split, which carries IQ samples, is the binding constraint when fronthaul capacity shrinks, not the downlink split. A sympathetic reader would take the paper as supplying the first real-network confirmation that these theoretical fronthaul formulas and the reconfiguration strategy carry over to commercial equipment.

What carries the argument

The load-bearing objects are the two linear fronthaul-rate identities $R^{ID}_{FH}=N_{RB}N_{SC}N_{MIMO}Q_M/T_S$ and $R^{IU}_{FH}=N_{RB}N_{SC}N_{MIMO}N_{IQ}/T_S$, where $N_{RB}$ is the number of active resource blocks, $N_{SC}=12$ subcarriers per block, $N_{MIMO}$ the number of active MIMO layers, $Q_M$ the modulation bits per symbol, $N_{IQ}$ the IQ-sample bit width, and $T_S$ the OFDM symbol duration. These formulas translate a radio-access configuration directly into a fronthaul capacity requirement, making cell reconfiguration—reducing $N_{RB}$ and $N_{MIMO}$—a way to match a degraded wireless link. The experiment's other essential mechanism is the variable-rate E-band transceiver pair inserted into the optical fronthaul and connected by a waveguide, whose capacity settings emulate the capacity reductions that rain or congestion would cause on a real wireless fronthaul link.

What would settle it

An outdoor test with a real E-band or millimeter-wave fronthaul link subjected to natural or artificial rain, while applying the same cell reconfiguration and measuring fronthaul rates; if the capacity at which the link fails to carry the configured cell, or the scaling of fronthaul rate with resource blocks and MIMO layers, deviates systematically from the waveguide-based measurements, the model transfer is refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that the fronthaul rate model of Eqs. 1 and 2 accurately describes a real commercial 5G fronthaul: downlink split ID carries raw data bits at a rate $R^{ID}_{FH}=N_{RB}N_{SC}N_{MIMO}Q_M/T_S$, while uplink split IU carries bit-encoded IQ symbols at $R^{IU}_{FH}=N_{RB}N_{SC}N_{MIMO}N_{IQ}/T_S$. Measured downlink fronthaul rates track Eq. 1 with only a small offset from antenna-control data; the uplink rate is several times higher than the downlink for the same access traffic because $N_{IQ}$ exceeds $Q_M$, and it comes in two regimes—a steep rise from antenna-control overhead at low load, then a slower linear rise as loaded OFDM symbols fill. The reconfiguration experiment shows an on/off threshold: at a given fronthaul capacity, the cell either delivers full throughput or collapses, and cutting resource blocks from 200 MHz to 100 MHz lowers both thresholds by about a factor of two, exactly as the model predicts. The authors conclude that downlink split ID is efficient while uplink split IU is the critical limitation during capacity reductions.

Load-bearing premise

The experiment's load-bearing premise is that two E-band radios connected directly to each other with their capacity turned down in steps faithfully reproduce how a real wireless fronthaul link behaves when rain or congestion cuts its capacity; if those artificial capacity steps miss the dynamics of an actual millimeter-wave link, the measured thresholds and the validated reconfiguration strategy may not transfer to deployed systems.

Editorial extensions

If this is right

Operators using split ID for downlink can compute the fronthaul rate needed from the cell configuration alone, since measured rates track Eq. 1 within a small antenna-control offset.
Halving the number of resource blocks roughly halves both the achievable access rate and the required fronthaul capacity, giving a simple software-only fallback when a wireless fronthaul link degrades.
Uplink split IU, not downlink, sets the capacity threshold for a cell: in the tested configuration the uplink needed about 2650 Mbps while downlink needed 1240 Mbps for an 850 Mbps access rate, so uplink-aware scheduling and compression matter most for resilience.
The on/off threshold behavior means a reconfiguration trigger can be placed near the measured minimum capacity for each cell configuration, turning a capacity drop into a predictable loss of peak rate rather than a link failure.

Reading between the lines

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

The paper does not test dynamic adaptation, but its linear model suggests the same reconfiguration rule could be applied per scheduling interval, scaling MIMO layers and bandwidth in near-real time as E-band capacity fluctuates.
The waveguide-based emulation implies that rain fading can be modeled as discrete capacity steps; a natural next experiment would map specific rainfall rates to E-band capacity settings and measure how often reconfiguration must trigger in a real outdoor link.
Because the uplink model's accuracy depends on the unknown IQ bit width $N_{IQ}$ and on vendor compression, a deployment that exposes or standardizes $N_{IQ}$ would make Eq. 2 a closed-form predictor rather than a fitted curve.
The finding that partially loaded OFDM symbols must still be sent in full on the uplink suggests that bursty, symbol-sparse traffic patterns do not reduce fronthaul demand; reshaping uplink scheduling to pack data into fewer symbols could lower the uplink threshold the paper identifies as the bottleneck.

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.

Desk Editor's Note

read the letter

The real asset here is the measurement campaign: a commercial 5G mmWave testbed with direct fronthaul bitrate measurements, clean scaling behavior, and a useful threshold analysis. The model itself is not new—it is restated from the authors' prior work [1]—and the paper is honest about that. What is new is the data, and for the downlink the data largely support the model. The factor-of-two scaling when the bandwidth is halved is clean, and the observation that uplink fronthaul, not downlink, is the binding constraint is worth having. Section VI-B is the strongest part of the paper. The setup with E-band transceivers connected by a waveguide is a clever way to make controlled capacity steps, and the authors are appropriately candid about the antenna-control offset and about not knowing the exact NIQ bitwidth.

The soft spots are real but not fatal. The main one is the gap between the title/abstract and what the experiments actually show. The configurations are switched manually, the E-band capacity is set in discrete steps, and no transitory capacity reduction or controller-driven adaptation is ever generated. So the paper demonstrates static feasibility of cell reconfiguration, not the dynamic 'smart fronthauling management' advertised in the introduction. That is a framing problem, fixable by rewriting the claims. Second, there are no error bars, no repeated runs, and no raw data; on a single proprietary testbed that limits the strength of the validation. Third, the validation is against the authors' own model from [1], with no independent benchmark or third-party data. Fourth, the waveguide emulation says nothing about rain-fading dynamics or transient behavior; the measured thresholds may transfer to a real wireless link, but the dynamic behavior will not. The stress-test note is right on this point, though I would not call the gap load-bearing for the static claims.

Proportionately: the downlink scaling result is solid, the uplink bottleneck insight is plausible, and the experimental dataset is rare and useful for the C-RAN/fronthaul community. The overclaim is the main issue. I would send this to a serious referee rather than desk reject, but I would require the authors to reframe the contribution as static feasibility under controlled capacity steps and to release the measurement data. With that, this becomes a worthwhile experimental report. Without it, the paper promises more than it delivers.

Referee Report

4 major / 4 minor

Summary. This paper reports experimental fronthaul measurements from a private 5G mmWave testbed (HFCL) in which two E-band transceivers are inserted into the optical fronthaul to emulate variable link capacity. Two experiments are described: first, the E-band capacity is manually reduced to find the minimum fronthaul rate that supports two cell configurations and the resulting access throughput is recorded; second, iperf3 is used to generate controlled DL and UL access traffic while fronthaul rate is measured for three cell configurations. The authors claim that the results validate the theoretical fronthaul rate formulas Eqs. (1) and (2) and demonstrate the viability of a cell reconfiguration strategy for adapting to reduced fronthaul capacity.

Significance. If the claims are substantiated, the paper provides rare empirical data from commercial C-RAN hardware with split ID/IU, showing that DL fronthaul rate scales roughly linearly with resource blocks and that the UL split IU is the binding constraint when fronthaul capacity is reduced. This is valuable for operators considering wireless or shared fronthaul, because most existing studies are simulation-based or use open-source testbeds. The strengths are the use of commercial equipment, a real mmWave radio access interface, and direct fronthaul bitrate measurements. However, the significance is reduced by the static nature of the experiments, the lack of uncertainty quantification, and the absence of raw data, all of which limit the strength of the quantitative validation claims.

major comments (4)

[Section VI-A and Introduction (Section I)] The experiment in Section VI-A exercises only static configuration changes: the E-band transceiver capacities are manually set and the two cell configurations of Table III are switched by the experimenters, not by a controller reacting to a capacity drop. The abstract and Section I claim demonstration of a 'dynamic and adaptive mechanism' and 'transitory fronthaul capacity reductions,' but no transitory or time-varying reduction is generated and no automatic reconfiguration loop is run. The paper should either add a dynamic experiment or explicitly limit the claim to static feasibility of cell reconfiguration.
[Section VI-B, Eq. (2)] The claim that Eq. (2) is validated is not quantitatively supported because NIQ is never measured; Section VI-B states that 'It is difficult to estimate the exact value of NIQ as it depends on dynamic compression algorithms.' The factor-of-two difference between configurations 2 and 3 in the UL curves of Figure 4a tests only the linear dependence on NMIMO, and the similar rates for configurations 1 and 3 test only the product NRB·NMIMO. To claim validation of Eq. (2), the authors need to provide an NIQ estimate (or at least bounds) and compare absolute rates; otherwise the conclusion should be restricted to the scaling behavior.
[Section VI-A, Figure 3] The reported thresholds do not fully support the 'reduction very close to a factor two' statement. Configuration 1 has thresholds of 3.1 Gbps DL and 2.7 Gbps UL, while configuration 2 is reported as 'approximately 1.6 Gbps in both directions'; this gives a DL ratio of about 1.94 but an UL ratio of only about 1.69, and the paper offers no error bars or repeated runs to distinguish the deviation from measurement scatter. The paper should provide repeated measurements and quantify the uncertainty before drawing quantitative agreement conclusions.
[Section V-A] The E-band emulation uses two transceivers connected by a waveguide and manually set capacity levels, rather than a free-space link subject to rain fading, adaptive modulation, or transient outage events. The paper acknowledges the waveguide but still describes the facility as capable of 'precisely replicating typical adverse atmospheric effects' and the experiment as emulating 'wireless links' capacity variations.' At minimum, the limitations of this emulation for drawing conclusions about real wireless fronthaul dynamics should be stated, since the measured thresholds and reconfiguration results may not transfer to a deployed link with continuous capacity fluctuations.

minor comments (4)

[Sections II and VI] Typos should be corrected: 'decisio' in Section II, 'he capacity' and 'TThis' in Section VI, and 'allof' in Section VI-B.
[Introduction] The section-cross-reference style is inconsistent: the introduction refers to 'section II discussed' and then to 'Section IV' without a clear ordering; make the forward references accurate and consistent.
[Figures 3 and 4] The figures are referenced in the text, but the axis labels, units, and the meaning of the black model line in Figure 4 are not described in the body; add explicit axis labels and a legend so the plots are self-contained.
[References] Reference [1] is an unreviewed arXiv preprint by the same authors; the paper should state the peer-review status or cite the published version if one exists.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the model and the measurements are independent, with no fitted parameter or self-citation chain forcing the claimed conclusions.

full rationale

The paper's derivation chain is empirical rather than circular. Equations 1 and 2 are recalled from the authors' prior work [1] as hypothesized formulas, and the experiments test those formulas against directly measured fronthaul bitrates. No parameter is fitted to the target measurements: the downlink comparison uses the known configured values of NRB, NMIMO, and QM, while the uplink comparison is qualitative because the paper explicitly states that NIQ is difficult to estimate and depends on dynamic compression algorithms. The cell-reconfiguration thresholds in Figure 3 are measured by sweeping E-band transceiver capacity and observing access throughput; they are not computed from the model and then presented as predictions. The observed halving of fronthaul requirement when RBs are halved is a measured trend compared with Eqs. 1 and 2, not a value forced by those equations. The self-citation [1] supplies the model and strategy under test, but the current paper's own experimental data constitute external evidence for them, so the citation is not load-bearing in a circular way. The waveguide-connected E-band link and the static, manually switched configurations are fidelity limitations for the claimed dynamic management viability, but they are not circularity: they concern whether the experiments support the generalization to dynamic wireless fronthaul behavior, not whether the reasoning assumes its conclusion.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new entities. It relies on the authors' own prior fronthaul rate model [1], on the assumption that a waveguide-coupled E-band pair emulates a real wireless link, and on an unmeasured antenna-control overhead that is used to absorb deviations. No parameters are fitted to data, but the unknown NIQ bitwidth leaves the uplink validation partially unconstrained.

free parameters (1)

NIQ (IQ sample bitwidth) = not estimated
Eq. 2 for the uplink fronthaul rate depends on NIQ, the number of bits per IQ sample. Section VI-B states 'It is difficult to estimate the exact value of NIQ as it depends on dynamic compression algorithms', and no value is reported, leaving the uplink validation partially unconstrained.

assumptions (4)

domain assumption Equations 1 and 2 from prior work [1] correctly model fronthaul rate requirements.
Section IV recalls these formulas without proof and uses them as the reference for the experimental validation; the same authors derived them in [1].
domain assumption The E-band transceiver pair connected by a waveguide faithfully emulates a wireless fronthaul link with variable capacity.
Section V-A replaces a free-space link with a waveguide 'to better confine and control the experiment' and assumes this reproduces the relevant capacity dynamics of a real wireless fronthaul.
domain assumption Antenna control overhead is minor in the testbed because hybrid beamforming is used.
Section IV states the overhead is expected to be minor and then ignores it; later, deviations from the model are attributed to this unmeasured overhead.
domain assumption Partially loaded OFDM symbols must be fully transmitted on the uplink fronthaul, and only fully empty symbols can be omitted.
Section VI-B uses this implementation behavior to explain the shape of the uplink fronthaul rate curves.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Towards Smart Fronthauling Management: Experimental Insights from a 5G Testbed." pith.science (2026). https://pith.science/paper/HWGME5WV

@misc{pith2026241113989,
  author       = {Pith},
  title        = {Pith review of: Towards Smart Fronthauling Management: Experimental Insights from a 5G Testbed},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HWGME5WV}},
  note         = {Machine review of arXiv:2411.13989}
}

read the original abstract

The fronthaul connection is a key component of Centralized RAN (C-RAN) architectures, consistently required to handle high capacity demands. However, this critical feature is at risk when the transport link relies on wireless technology. Fortunately, solutions exist to enhance the reliability of wireless links. In this paper, we recall the theoretical fronthaul model, present a dynamic reconfiguration strategy and perform a conclusive experiment. Specifically, we showcase the setup of a wireless fronthaul testbed and discuss the resulting measurements. For this task, we leveraged the commercial hardware provided by the High-Frequency Campus Lab (HFCL), a private 5G network with millimeter wave (mmWave) radio access interface. Our experiments provide original data on the fronthaul utilization in this real deployment, demonstrating both a good accordance with the theoretical model discussed in [1] and the viability of one stabilizing solution.

Figures

Figures reproduced from arXiv: 2411.13989 by the authors.

**Figure 2.** The HFCL includes an Active Antenna Unit (AAU) [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 2.** High-Frequency Campus Lab RAN equipment involved in the experiment [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Figures from the paper (2 more)

**Figure 3.** Figure 3: Impact of Cell Reconfiguration on fronthaul rates [PITH_FULL_IMAGE:figures/full_fig_p004_3.png]

**Figure 4.** Figure 4: Relation between fronthaul rate and access rate [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

16 extracted references · 15 canonical work pages

[1]

Shaping Radio Access to Match Variable Wireless Fronthaul Quality in Next-Generation Networks

M. Morini, E. Moro, I. Filippini, D. D. Donno, and A. Capone, “Shaping radio access to match variable wireless fronthaul quality in next-generation networks.” https://arxiv.org/abs/2406.15899, 2024

work page Pith review arXiv 2024
[2]

Wireless fronthaul for 5g and future radio access networks: Challenges and enabling technologies,

M. Jiang, J. Cezanne, A. Sampath, O. Shental, Q. Wu, O. Koymen, A. Bedewy, and J. Li, “Wireless fronthaul for 5g and future radio access networks: Challenges and enabling technologies,” IEEE Wireless Communications, vol. 29, no. 2, pp. 108–114, 2022

work page 2022
[3]

Fronthaul design for wireless networks,

I. Sousa, N. Sousa, M. P. Queluz, and A. Rodrigues, “Fronthaul design for wireless networks,” Applied Sciences, vol. 10, no. 14, 2020

work page 2020
[4]

eCPRI Specification V2.0,

Ericsson AB, Huawei Technologies Co. Ltd, NEC Corporation and Nokia, “eCPRI Specification V2.0,” tech. rep., 2019-05-10

work page 2019
[5]

Control, User and Synchronization Plane Specification, 0-R003- v13.00,

O-RAN Alliance, Working Group 4 (Open Fronthaul Interfaces WG), “Control, User and Synchronization Plane Specification, 0-R003- v13.00,” tech. rep., 2023

work page 2023
[6]

A survey of the functional splits proposed for 5g mobile crosshaul networks,

L. M. P. Larsen, A. Checko, and H. L. Christiansen, “A survey of the functional splits proposed for 5g mobile crosshaul networks,” IEEE Communications Surveys & Tutorials, vol. 21, no. 1, pp. 146–172, 2019

work page 2019
[7]

Darts - an analysis tool for disaggregated ran transport,

Small Cell Forum, “Darts - an analysis tool for disaggregated ran transport,” tech. rep., 2023

work page 2023
[8]

5g ran: Functional split orchestration optimization,

S. Matoussi, I. Fajjari, S. Costanzo, N. Aitsaadi, and R. Langar, “5g ran: Functional split orchestration optimization,” IEEE Journal on Selected Areas in Communications , vol. 38, no. 7, pp. 1448–1463, 2020

2020

Show all 16 references

[9]

Dynamic functional split adapta- tion in next-generation radio access networks,

A. Mart ´ınez Alba and W. Kellerer, “Dynamic functional split adapta- tion in next-generation radio access networks,” IEEE Transactions on Network and Service Management, vol. 19, no. 3, pp. 3239–3263, 2022

2022
[10]

Fronthaul-aware software- defined joint resource allocation and user scheduling for 5g networks,

C.-F. Liu, S. Samarakoon, and M. Bennis, “Fronthaul-aware software- defined joint resource allocation and user scheduling for 5g networks,” in 2016 IEEE Globecom Workshops (GC Wkshps) , pp. 1–6, 2016

2016
[11]

Enabling flexible link capacity for ecpri-based fronthaul with load-adaptive quantization resolution,

L. Li, M. Bi, H. Xin, Y . Zhang, Y . Fu, X. Miao, A. M. Mikaeil, and W. Hu, “Enabling flexible link capacity for ecpri-based fronthaul with load-adaptive quantization resolution,” IEEE Access, vol. 7, pp. 102174– 102185, 2019

2019
[12]

On the feasibility of mac and phy split in cloud ran,

G. Mountaser, M. L. Rosas, T. Mahmoodi, and M. Dohler, “On the feasibility of mac and phy split in cloud ran,” in 2017 IEEE Wireless Communications and Networking Conference (WCNC) , pp. 1–6, 2017

2017
[13]

Experimental oai-based testbed for evaluating the impact of different functional splits on c-ran per- formance,

A. I. Salama and M. M. Elmesalawy, “Experimental oai-based testbed for evaluating the impact of different functional splits on c-ran per- formance,” in 2019 Novel Intelligent and Leading Emerging Sciences Conference (NILES), vol. 1, pp. 170–173, 2019

2019
[14]

Challenges and opportunities in wireless fronthaul,

D. Townend, R. Husbands, et al. , “Challenges and opportunities in wireless fronthaul,” IEEE Access, vol. 11, pp. 106607–106619, 2023

2023
[15]

NR; User Equipment (UE) radio access capabilities,

3GPP, “NR; User Equipment (UE) radio access capabilities,” Technical Specification (TS) 38.306, 3GPP, 03 2023. Version 17.4.0

2023
[16]

NR; Base Station (BS) radio transmission and reception (Section 5.3.2),

3GPP, “NR; Base Station (BS) radio transmission and reception (Section 5.3.2),” Technical Specification (TS) 38.104, 3GPP, 02 2024. Version 17.12.0

2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Shaping Radio Access to Match Variable Wireless Fronthaul Quality in Next-Generation Networks

M. Morini, E. Moro, I. Filippini, D. D. Donno, and A. Capone, “Shaping radio access to match variable wireless fronthaul quality in next-generation networks.” https://arxiv.org/abs/2406.15899, 2024

work page Pith review arXiv 2024

[2] [2]

Wireless fronthaul for 5g and future radio access networks: Challenges and enabling technologies,

M. Jiang, J. Cezanne, A. Sampath, O. Shental, Q. Wu, O. Koymen, A. Bedewy, and J. Li, “Wireless fronthaul for 5g and future radio access networks: Challenges and enabling technologies,” IEEE Wireless Communications, vol. 29, no. 2, pp. 108–114, 2022

work page 2022

[3] [3]

Fronthaul design for wireless networks,

I. Sousa, N. Sousa, M. P. Queluz, and A. Rodrigues, “Fronthaul design for wireless networks,” Applied Sciences, vol. 10, no. 14, 2020

work page 2020

[4] [4]

eCPRI Specification V2.0,

Ericsson AB, Huawei Technologies Co. Ltd, NEC Corporation and Nokia, “eCPRI Specification V2.0,” tech. rep., 2019-05-10

work page 2019

[5] [5]

Control, User and Synchronization Plane Specification, 0-R003- v13.00,

O-RAN Alliance, Working Group 4 (Open Fronthaul Interfaces WG), “Control, User and Synchronization Plane Specification, 0-R003- v13.00,” tech. rep., 2023

work page 2023

[6] [6]

A survey of the functional splits proposed for 5g mobile crosshaul networks,

L. M. P. Larsen, A. Checko, and H. L. Christiansen, “A survey of the functional splits proposed for 5g mobile crosshaul networks,” IEEE Communications Surveys & Tutorials, vol. 21, no. 1, pp. 146–172, 2019

work page 2019

[7] [7]

Darts - an analysis tool for disaggregated ran transport,

Small Cell Forum, “Darts - an analysis tool for disaggregated ran transport,” tech. rep., 2023

work page 2023

[8] [8]

5g ran: Functional split orchestration optimization,

S. Matoussi, I. Fajjari, S. Costanzo, N. Aitsaadi, and R. Langar, “5g ran: Functional split orchestration optimization,” IEEE Journal on Selected Areas in Communications , vol. 38, no. 7, pp. 1448–1463, 2020

2020