Pith. sign in

REVIEW 4 major objections 5 minor 19 references

Quality over Quantity: Value-Driven Distributed Congestion Control for the Collective Perception Service

T0 review · 4 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read An adaptive Value-of-Information threshold at the facilities layer transmits a greater share of high-value sensed objects than the incumbent quantity-based selector, under both homogeneous and heterogeneous network conditions, while holding

desk verdict Adaptive VoI thresholding is a real idea, but the evaluation doesn't enforce the FCL byte budget for the quality selector, so the headline gains may just be extra bytes. read the letter →

arxiv 2607.18495 v1 pith:MSOCN6AH submitted 2026-07-20 cs.NI

Tengfei Lyu , Florian A. Schiegg , Md Noor-A-Rahim , Dirk Pesch , Aisling O'Driscoll This is my paper · ORCID
classification cs.NI
keywords CollectivePerceptionServiceDistributedCongestionControlValueofInformationV2XSensorDataSharingAdaptivethresholdingContentselectionChannelBusyRatio
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 is trying to show that congestion control for the collective perception service in vehicle-to-everything networks should select which sensed objects to transmit based on their value, not just fill each message up to a bit budget. It proposes an adaptive VoI threshold that rises when the local object set exceeds the allocated budget and falls when there is spare capacity, so low-value objects are dropped first as congestion worsens. The paper compares this 'quality' selector to the standard 'quantity' selector under simulated homogeneous and heterogeneous conditions, and reports that it transmits a larger share of the total generated VoI while keeping channel load close to target. If true, this is a practical improvement: it would deliver more perception-critical information to nearby vehicles under the same radio constraints.

What carries the argument

The adaptive VoI threshold, updated via Equations 5–6 as a proportional controller on the ratio of total sensed content bytes to the allocated Facilities-layer byte budget, then applied in Equation 7 to filter objects before CPM assembly. This threshold does the work of converting a fixed bit budget into a variable selectivity that responds to local content load and value distribution, rather than simply filling messages with the highest-ranked objects.

What would settle it

Run the same comparison with a concrete, field-calibrated VoI function and temporally correlated object values. If the adaptive threshold no longer retains a higher share of network transmitted VoI at matched CBR, or if it causes channel-load oscillations, the central claim fails. A minimal check: if all objects have nearly equal VoI, the threshold cannot discriminate and the reported gains should collapse.

Watch

Extended reading notes

Core claim

On its own terms, the central claim is that an adaptive Value-of-Information threshold at the facilities layer of the ITS protocol stack, operating on top of an A-DCC-like rate controller, retains a higher network-transmitted VoI (the fraction of generated object value that actually gets transmitted across all stations) than the incumbent quantity-based selector, particularly when stations differ in how many objects they see and in the spread of object values. Equations 5–7 implement this: compute the load ratio ρ = total content size / allocated byte budget; if ρ > 1 raise the threshold by a clipped proportional step, if ρ < 1 lower it; then transmit all objects with VoI ≥ θ. The simulation

Load-bearing premise

The paper assumes every station can compute a trustworthy Value-of-Information score for every sensed object in real time; the paper itself notes the VoI function is 'yet to be defined', so the simulation samples scores from synthetic distributions rather than a real perception pipeline.

Editorial extensions

If this is right

  • Under the same channel-load target, a higher fraction of the total generated object value is received by the network, so vehicles get more of the perception-critical information without increasing channel occupancy.
  • The benefit grows with heterogeneity: stations in sparse, medium and dense perception profiles get different effective thresholds, as seen in the three-peaked threshold distribution under heterogeneous conditions.
  • The quantity selector admits low-VoI objects when higher-value content is present in other stations; the quality selector avoids this, so scarce budget is directed to high-VoI content.
  • CBR remains near the target under both selectors, so the VoI gain is not bought by sacrificing congestion control.
  • The approach is compatible with the reference DCC-FAC architecture and only changes how the CPS uses the allocated budget.

Reading between the lines

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

  • If validated with a real VoI pipeline, the same thresholding idea could be extended upward to the bandwidth management entity: instead of equal per-station budgets, the budget itself could be made VoI-dependent, something the paper explicitly leaves for future work.
  • The adaptive threshold is a control loop that could be analyzed for stability and convergence with real VoI distributions; the simulation only samples independent scores per event, so temporal correlation of object values (e.g., a newly seen obstacle staying valuable for several hundred ms) might change how the threshold oscillates.
  • A testable extension: apply the same quality selector to non-CPM services or mixed service classes where the bit budget is shared, since the threshold mechanism only requires a per-object value and a byte budget.
  • The 'boundary vs threshold' distinction suggests a metric: if a quantity selector's effective boundary is often lower than a quality selector's threshold at the same load, the quality selector is discarding low-value objects that the quantity selector would transmit; this could be measured directly in field trials.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a Facilities-layer congestion-control content selector for the Collective Perception Service (CPS) that filters perceived objects/regions by an adaptive Value of Information (VoI) threshold. The threshold is updated per CPM generation event using the ratio of total sensed content size to the allocated Facilities-layer bit budget (Eqs. 4-6), and all content above the threshold is included in the CPM (Eq. 7). The method is evaluated against the ETSI DCC FAC quantity selector in a custom simulation with synthetic VoI distributions, under homogeneous and heterogeneous network knowledge and with 5, 10, and 15 ITS-S. The headline claim is that the proposed quality selector achieves higher “network transmitted VoI” (Eq. 11) while keeping CBR near the target, with gains of 6.5-21.3 percentage points (Table II).

Significance. If the claims hold, the paper makes a useful, standards-aligned contribution: it introduces a simple, distributed, content-value-aware mechanism that can be implemented at the Facilities layer without modifying the ETSI DCC FAC reference architecture. The focus on heterogeneous network knowledge and on moving beyond quantity-based selection addresses a real limitation of the current ETSI approach. The paper is clearly written and the simulation design is transparent about its synthetic nature. However, the central evaluation currently rests on two assumptions that the authors explicitly flag: the VoI function is “yet to be defined” (Sec. II-C) and is instead sampled from arbitrary distributions (Table I), and the resource-limit enforcement of Eq. (7) is not demonstrated. The main quantitative claim therefore needs additional support before the results can be considered established.

major comments (4)
  1. [Sec. III-C, Eq. (7)] The selector C_j(t) = {i in O_j(t) : VoI_i >= theta_j(t)} includes every object above the threshold without any post-selection check that the resulting CPM fits the Facilities-layer bit budget B(t) from Eq. (2). The load ratio rho_j(t) in Eq. (4) is computed from L_j(t), the total size of all sensed content, not from the size of the selected set. A single adaptive-threshold update cannot guarantee that the sum of the sizes of the selected objects is <= B(t); in dense scenes with many high-VoI objects, the CPM can exceed the allocated budget. The ETSI quantity selector explicitly ranks content and truncates at B(t), so the comparison in Table II is not “under a common allocated Facilities layer bit budget” unless the quality selector also enforces the cap. The paper’s own admission in Sec. V-C that “occasional over-utilisation of the channel can occur” and the attribution of this to the B
  2. [Sec. II-C / Sec. IV, Table I] The VoI function is explicitly “yet to be defined” (Sec. II-C), and the simulation replaces it with synthetic samples drawn from the distributions in Table I. This means the whole evaluation is carried out under a hypothetical VoI scoring pipeline. The proposed threshold dynamics may behave very differently when VoI scores are correlated with object properties, spatial relationships, or prior awareness, as in the real VoI models cited in [15] and [17]. The paper acknowledges this in the future-work section, but the central claim — that the adaptive threshold “consistently retains higher network transmitted VoI” — is not yet supported for any concrete VoI model. At minimum, the evaluation should be repeated with at least one published VoI function, or a sensitivity analysis over plausible VoI distributions should be provided.
  3. [Sec. V-A, Eq. (11)] The metric P in Eq. (11) is exactly the objective that the proposed selector is designed to maximize: raising the VoI threshold and transmitting only high-VoI objects mechanically increases the numerator (transmitted VoI) when the total generated VoI is held fixed. Consequently, the gains in Table II are partly true by construction. The non-tautological part of the claim is whether the quality selector achieves this under equal resource consumption and whether the CBR remains near the target. Because the budget-compliance issue above is unresolved, the reported advantage may simply reflect that the quality selector transmits more bytes. The authors should report the total transmitted bytes (or the average CPM size) for both selectors and, ideally, normalize P by the transmitted bytes to demonstrate that the gain is due to better selection rather than higher channel usage.
  4. [Table II / Figs. 4-5] The main quantitative results in Table II are reported as single point estimates with no confidence intervals, no number of simulation runs, and no statistical tests. Figs. 4 and 5 likewise show distributions but without error bars or run counts. Given that the VoI values and content-set sizes are drawn stochastically, the differences between the two selectors could be within run-to-run variability, particularly for the small 5-ITS-S cases (e.g., homogeneous gain of 6.53 percentage points). The claim that the proposed method “consistently retains” higher network transmitted VoI needs at least standard errors or box plots over repeated independent runs.
minor comments (5)
  1. [Abstract and Sec. I] The abstract and introduction state that the method is benchmarked against “state of the art approaches”, but the evaluation (Sec. V) compares only against the ETSI DCC FAC quantity selector. The recent value-based methods by Wolff et al. [15] and Sepulcre et al. [16] are discussed but not included in the benchmark. Either implement those baselines or rephrase the claim as a comparison against the ETSI quantity selector.
  2. [Sec. III-C, Eq. (10)] Eq. (10) is typeset with stray “&” characters and is difficult to parse. The relationship between the CPM overhead H in Eq. (2) and the size model in Eq. (10) should be made explicit. Also, the term “(68±20)” is described as a truncated Gaussian with standard deviation 10 clipped to [48,88], but “±20” suggests a different range; please clarify.
  3. [Various] There are several minor language issues: “it’s Value of Information” in the abstract, “This is largely depends” in Sec. III-B, and inconsistent spelling of “homogeneous”/homogenous”. These do not affect the technical content but should be corrected.
  4. [Sec. IV] The simulation assumes that object VoI values are sampled independently at each CPM generation event. This ignores temporal correlation and spatial correlation across objects and ITS-S. While this is acceptable for a first controlled study, the authors should state this as a limitation more prominently, since it may bias the results toward favoring a threshold-based selector.
  5. [Fig. 2 and Fig. 3] The “?” symbols in the figure captions (likely intended as the value of P) are not defined in the caption text. Also, the captions report percentages and numbers in parentheses (e.g., “69.23% (7.65)”) without explaining what the parenthetical value represents. Please clarify.

Circularity Check

1 steps flagged · score 6.0 of 10

VoI gain is not tested under the stated bit budget: Eq. 7 admits all above-threshold content with no cap at B(t), so part of the Table II gain is built into the selection rule.

  1. other [Sec. III-C, Eq. (7); Sec. V-A, Eq. (11); Sec. V-A Table II]
    "Cj(t) ={i∈ Oj(t) : VoI i ≥θj(t)}. (7) Hence, all content whose V oI is above the current threshold is included in the CPM. ... P = Σ VoI_tx / Σ VoI_gen (11) ... It confirms that given the same Facilities layer bit budget, it is beneficial to regulate object eligibility through an adaptive V oI threshold."

    Eq. 7 selects every object above θ_j(t) and does not truncate the selected set to the per-CPM budget B(t) of Eq. 2. The headline metric P is the ratio of transmitted to generated VoI, so permitting extra above-threshold objects to be transmitted increases P by construction. The ETSI quantity baseline is explicitly truncated at B(t), so the comparison in Table II is not 'under a common allocated Facilities layer bit budget' unless the quality selector also enforces the cap. The paper's own Sec. V-C concedes 'occasional over-utilisation of the channel can occur,' which is consistent with the selector sending more bytes than allocated; the VoI gain is therefore partly an artifact of relaxing the resource constraint rather than an independent demonstration of smarter selection.

full rationale

The proposed adaptive-threshold controller itself is not circular: θ_j(t) is updated from the load ratio ρ_j(t)=L_j(t)/B(t), and whether that closed loop keeps CBR near target is an empirical question answered by simulation. The paper's self-citations (e.g., [17]) are not load-bearing; they merely illustrate possible VoI functions. However, the central comparative claim—higher network-transmitted VoI under the same allocated bit budget—is compromised by the absence of any budget cap in the quality selector's selection rule. Because Eq. 7 includes all content above the threshold regardless of B(t), and because the metric P rewards transmitting more high-VoI content, a substantial part of the reported gain is a consequence of the selector's construction rather than of the threshold adaptation. The CBR results provide some independent evidence of congestion control, but the paper does not show that the quality selector respects the per-CPM byte budget used for the quantity baseline, and it concedes over-utilisation. This warrants a partial-circularity score of 6 rather than a higher score, since the adaptive-threshold mechanism itself has independent content and the simulation is reproducible, while the VoI function is admittedly 'yet to be defined' and synthetic VoI distributions are used.

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

The mechanism's behavior is imported from assumptions about VoI availability, an abstract CBR mapping, equal BME allocation, and implicit FCL compliance. The headline comparison also depends on hand-chosen synthetic profiles in Table I. No new physical entities are postulated.

free parameters (5)
  • Adaptive threshold controller gains
    beta_theta, alpha_theta, Delta_theta_min, Delta_theta_max, theta_min, theta_max in Eqs. 5-6 are not given in the paper; their values control selectivity and are not justified or swept.
  • A-DCC rate-control coefficients alpha, beta
    Used in the global delta update (Eq. 1-style controller); values are not reported, and the equilibrium CBR depends on them.
  • CPS/FCL budget constants R, H, TCPM, R_abs, s_f
    Simulation constants in Eqs. 2 and 8 set the per-message bit budget and the CBR mapping; they are not specified and control how many objects can fit.
  • Object size model = mean 68 bits, std 10, clipped to [48,88]; header 127 bytes
    Eq. 10 uses a truncated Gaussian for per-object size; chosen by hand, this affects whether low-VoI objects crowd out high-VoI ones.
  • Network-knowledge profiles = see Table I: mu/sigma for object count and VoI per profile
    Homogeneous and heterogeneous profiles are synthetic inputs chosen to demonstrate the mechanism; the numerical gains depend directly on these distributions.
assumptions (4)
  • domain assumption Per-object VoI scores are available and accurate at each ITS-S
    The selector (Eq. 7) and the evaluation (Table I) assume VoI is known; the paper states the VoI function is yet to be defined (Sec. II-C, Sec. III-C).
  • domain assumption The abstract channel-load model with scaling factor s_f (Eq. 8) approximates real ITS-G5 congestion
    CBR is computed from total bytes only; no PHY/MAC fading, collisions, or packet errors are modeled. The central claim about maintaining target CBR rests on this simplification.
  • domain assumption The BME allocates equal Facilities-layer bit budgets to all ITS-S
    Section V.C states the A-DCC type controller allocates resources equally across ITS-S; the comparison inherits this and no content-aware BME allocation is considered.
  • domain assumption The selected content set after thresholding conforms to the FCL/MHE resource limit
    Eq. 7 does not cap C_j(t) by B(t); the paper relies on the threshold and global delta to keep load near target but does not explicitly model MHE enforcement of non-compliant messages.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quality over Quantity: Value-Driven Distributed Congestion Control for the Collective Perception Service." pith.science (2026). https://pith.science/paper/MSOCN6AH

@misc{pith2026260718495,
  author       = {Pith},
  title        = {Pith review of: Quality over Quantity: Value-Driven Distributed Congestion Control for the Collective Perception Service},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MSOCN6AH}},
  note         = {Machine review of arXiv:2607.18495}
}
read the original abstract

While the Collective Perception Service (CPS) enables the exchange of sensor information among Intelligent Transport System Stations (ITS-S'), frequent transmission of Collective Perception Messages (CPMs), their highly variable size, and load from other vehicular services can cause severe channel congestion. Existing Distributed Congestion Control (DCC) Access layer mechanisms typically regulate channel load without considering the relative importance of the objects carried in CPMs. This limits their ability to preserve high-value information under constrained radio resources. More recently, Facilities layer DCC mechanisms attempt to prioritise high value objects within the specified radio resource limits but may not operate well in heterogeneous environments where the number of sensed objects and their importance can vary significantly over time or between ITS-S'. This paper proposes a value-based DCC Facilities layer 'quality' selector that couples a Value of Information (VoI) per bit rate controller with object-level selection. It is benchmarked against state of the art approaches from standards and the literature, with results showing that the proposed method maintains channel load near the target CBR while retaining more high-VoI objects than state of the art approaches, thereby improving the dissemination of perception-critical information.

Figures

Figures reproduced from arXiv: 2607.18495 by the authors.

Figure 1
Figure 1. The ETSI DCC Facilities layer (DCC-FAC) Architec [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Network Transmitted VoI (P) of three DCC approaches under homogeneous network knowledge, where the number of perceived objects/regions and the overall VoI distribution are statistically similar across ITS-S’. H H H 0.82 0.69 0.53 0.48 0.43 0.41 0.36 0.12 0.09 0.98 0.98 0.98 0.97 0.96 0.94 0.94 0.92 0.91 0.91 0.72 0.56 0.39 0.39 0.21 0.12 0.10 0.90 0.90 0.90 0.89 0.56 0.41 0.23 ? = 59.66% (12.35) (a) ETSI DCC FAC qua… view at source ↗
Figure 3
Figure 3. Network Transmitted VoI (P) of the quantity-based and proposed quality content DCC content selectors under heterogeneous network knowledge, where the number of perceived objects/regions and the distribution of VoI values may differ across ITS-S’. in Fig. 2c introduces a stronger form of value-awareness. It applies an adaptive minimum VoI threshold so that content with insufficient value is discounted before packet a… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Histograms of generated VoI with the objects included for transmission shown in blue. This assumes [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Representative CBR violin plots for the 15 ITS-S’ case under (a) homogeneous and (b) heterogeneous network knowledge. The dashed line indicates the target CBR. The omitted 5 and 10 ITS-S’ cases exhibit the same behaviour. study assumes a fixed CPM generation interval o…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 1 linked inside Pith

  1. [15]

    Uncertainty and prioritization: Empirical evaluation of a voi-based cpm generation pipeline using real-world data,

    V . A. Wolff, E. Xhoxhi, F. A. Schiegg, and J.-E. Neumann, “Uncertainty and prioritization: Empirical evaluation of a voi-based cpm generation pipeline using real-world data,” inIEEE Vehicular Networking Confer- ence (VNC), 2025, pp. 1–8

    2025

  2. [17]

    Accuracy and relevance: A value of information based prioritisation of perceived objects for the etsi collective perception service,

    T. Lyu, S. Li, F. A. Schiegg, M. Noor-A-Rahim, D. Pesch, and A. O’Driscoll, “Accuracy and relevance: A value of information based prioritisation of perceived objects for the etsi collective perception service,” inIEEE Vehicular Networking Conference (VNC), 2025, pp. 1–8

    2025

  3. [1]

    Intelligent Transport System (ITS); Vehicular Communica- tions,

    ETSI ITS, “Intelligent Transport System (ITS); Vehicular Communica- tions,”Basic Set of Applications; Collective Perception Service; Release 2 TS 103 324 V2.1.1, 2023

    2023

  4. [2]

    A first study on the spectrum needs for release 2 V2X services,

    E. Xhoxhi and F. A. Schiegg, “A first study on the spectrum needs for release 2 V2X services,” inIEEE 98th Vehicular Technology Conference (VTC2023-Fall), 2023, pp. 1–6

    2023

  5. [3]

    Enhancing vehicular network efficiency: the impact of object data inclusion in the collective perception service,

    A. Figueiredo, P. Rito, M. Lu ´ıs, and S. Sargento, “Enhancing vehicular network efficiency: the impact of object data inclusion in the collective perception service,”IEEE Open Journal of Intelligent Transportation Systems, 2024

    2024

  6. [4]

    Intelligent Transport Systems (ITS); Cross Layer DCC Management Entity for operation in the ITS G5A and ITS G5B medium,

    ETSI ITS, “Intelligent Transport Systems (ITS); Cross Layer DCC Management Entity for operation in the ITS G5A and ITS G5B medium,”TS 103 175 V1.1.1, 2015

    2015

  7. [5]

    Limeric: A linear adaptive message rate algorithm for dsrc congestion control,

    G. Bansal, J. B. Kenney, and C. E. Rohrs, “Limeric: A linear adaptive message rate algorithm for dsrc congestion control,”IEEE Transactions on Vehicular Technology, vol. 62, no. 9, pp. 4182–4197, 2013

    2013

  8. [6]

    Achieving weighted-fairness in message rate-based congestion control for dsrc systems,

    G. Bansal and J. B. Kenney, “Achieving weighted-fairness in message rate-based congestion control for dsrc systems,” inIEEE 5th Inter- national Symposium on Wireless Vehicular Communications (WiVeC), 2013, pp. 1–5

    2013

Show all 19 references
  1. [7]

    Embarc: Error model based adaptive rate control for vehicle-to-vehicle communica- tions,

    G. Bansal, H. Lu, J. B. Kenney, and C. Poellabauer, “Embarc: Error model based adaptive rate control for vehicle-to-vehicle communica- tions,” inProceeding of the tenth ACM international workshop on Vehicular inter-networking, systems, and applications, 2013, pp. 41–50

    2013

  2. [8]

    Swerc: Self-weighted semi-cooperative dsrc congestion control based on limeric,

    T. Lorenzen, “Swerc: Self-weighted semi-cooperative dsrc congestion control based on limeric,” inIEEE 86th Vehicular Technology Confer- ence (VTC-Fall), 2017, pp. 1–7

    2017

  3. [9]

    Intelligent Transport Systems (ITS); Facilities layer function; Communication congestion control; Release 2,

    ETSI ITS, “Intelligent Transport Systems (ITS); Facilities layer function; Communication congestion control; Release 2,”TS 103 141 V2.1.1, 2021

    2021

  4. [10]

    Intelligent Transport Systems (ITS); Facilities layer function; Multi-Channel Operation (MCO) for Cooperative ITS (C-ITS); Release 2,

    ETSI ITS, “Intelligent Transport Systems (ITS); Facilities layer function; Multi-Channel Operation (MCO) for Cooperative ITS (C-ITS); Release 2,”Basic Set of Applications TS 103 141 V2.2.1, 2022

    2022

  5. [11]

    Intelligent Transport System (ITS),

    ETSI ITS, “Intelligent Transport System (ITS),”Facilities layer; Radio Resource Management Study; Release 2; TR 104 073; V2.1.1, 2025

    2025

  6. [12]

    Intelligent transport systems (ITS); Decentralized Conges- tion Control Mechanisms for Intelligent Transport Systems operating in the 5 Ghz range; Access layer part,

    ETSI ITS, “Intelligent transport systems (ITS); Decentralized Conges- tion Control Mechanisms for Intelligent Transport Systems operating in the 5 Ghz range; Access layer part,”TS 102 687 V1.2.1, 2018

    2018

  7. [13]

    Enhancing ETSI DCC for multi- service vehicular safety communication,

    M. I. Khan, J. H ¨arri, and S. Sesia, “Enhancing ETSI DCC for multi- service vehicular safety communication,” inIEEE 92nd Vehicular Tech- nology Conference (VTC2020-Fall), 2020, pp. 1–5

    2020

  8. [14]

    In vehicle resource orchestration for multi-v2x services,

    M. I. Khan, S. Sesia, and J. Harri, “In vehicle resource orchestration for multi-v2x services,” inIEEE 90th Vehicular Technology Conference (VTC2019-Fall), 2019, pp. 1–5

    2019

  9. [16]

    Demand-and priority- aware adaptive congestion control for heterogeneous v2x service require- ments,

    M. Sepulcre, J. Tortosa-Garcia, and J. Gozalvez, “Demand-and priority- aware adaptive congestion control for heterogeneous v2x service require- ments,”arXiv preprint arXiv:2603.01134, 2026

    2026 arXiv

  10. [18]

    Improving object selection for collective perception messages under congestion,

    A. Figueiredo, J. Amaral, P. Rito, M. Lu ´ıs, and S. Sargento, “Improving object selection for collective perception messages under congestion,” Available at SSRN 5389372

  11. [19]

    V2X congestion control for multi-channel operation: a scalable validation in virtualized envi- ronments,

    M. Sepulcre, Y . Guadalcazar, M. A. Fornell, G. Thandavarayan, F. P. Vera, J. Gozalvez, and A. Mohammadisarab, “V2X congestion control for multi-channel operation: a scalable validation in virtualized envi- ronments,” inIEEE 101st Vehicular Technology Conference (VTC2025- Spri...

    2025

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.