Pith. sign in

REVIEW 4 major objections 5 minor 13 references

Comparison of STR and EMLSR Performance in Wi-Fi 7 MLO

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read STR mode consistently outperforms EMLSR in the simulated Wi-Fi 7 multi-link scenarios, at every MCS, bandwidth, and interference setting tested.

desk verdict A straightforward ns-3 STR-vs-EMLSR comparison whose central ranking holds but whose conclusion overstates consistency, since the paper's own access-delay plots favor EMLSR in several regimes. read the letter →

arxiv 2501.04149 v1 pith:UEJ4BKYT submitted 2025-01-07 cs.NI

Aishwarya Choorakuzhiyil , Kevin Ho , Sara Reyes This is my paper · ORCID
classification cs.NI
keywords Wi-Fi7IEEE802.11bemulti-linkoperationsimultaneoustransmitandreceiveenhancedsingleradiosaturatedthroughputMAClatencyinterference
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

This paper tries to establish that in Wi-Fi 7 multi-link operation, STR mode beats EMLSR mode across the network configurations it simulates. STR delivers roughly double the saturated throughput of EMLSR in the baseline case, about 120 Mbps versus 60 Mbps, and keeps queuing, access, and end-to-end delays lower because it uses both links for simultaneous transmission, while EMLSR is limited by a single radio that can transmit on only one link at a time. The authors also show that both multi-link modes outperform single-link operation, which saturates around 30 Mbps. They position EMLSR as better suited for power-sensitive devices, but the energy advantage is asserted rather than measured in the reported simulations.

What carries the argument

The central mechanism is the radio-constraint difference inside multi-link operation. STR devices have two radios and can transmit and receive simultaneously on both links; EMLSR devices have one radio that can listen on both links but must pick one link for transmission. The paper exercises this mechanism by varying the offered load $\lambda$ across $10^{-5}$ to $10^{-1}$, varying link 1's MCS over $\{2,4,6,8\}$, varying link 1's bandwidth over $\{20,40,80\}$ MHz, and adding single-link interferers symmetrically or asymmetrically, then comparing saturated throughput and mean queuing, access, and end-to-end delays. The recurring result is that EMLSR's single-radio switching constraint caps its throughput and raises delays, while STR's dual-link operation raises the saturation point and absorbs more offered traffic.

What would settle it

Set up a Wi-Fi 7 client that can be switched between STR and EMLSR, put it on a single-AP network with controlled offered load, and measure saturated throughput and queuing, access, and end-to-end delay under the same MCS, bandwidth, and interference conditions used here; if EMLSR's saturated throughput is not markedly below STR's, or if STR's delays are not lower, the central claim would be contradicted.

Watch

Extended reading notes

Core claim

The paper claims that in the simulated Wi-Fi 7 multi-link networks, simultaneous transmit and receive (STR) is consistently the better mode: it reaches a saturated throughput about twice that of enhanced multi-link single radio (EMLSR) in the base single-AP, five-station network, and it keeps queuing, access, and end-to-end delays lower across MCS, bandwidth, and interference variations. EMLSR, which listens on two links but transmits on only one, lands between STR and single-link operation, with the three modes saturating at roughly 120, 60, and 30 Mbps, respectively. The conclusion states flatly that STR consistently outperformed EMLSR, while noting that EMLSR has a design goal of energy efficiency that this study does not quantify.

Load-bearing premise

The ranking rests on the simulator's two mode implementations faithfully matching real 802.11be behavior, especially EMLSR's single-radio constraint and STR's simultaneous dual-link operation, and the paper does not validate the simulation against measurements.

Editorial extensions

If this is right

  • In the simulated base network, STR saturates at about 120 Mbps, EMLSR at about 60 Mbps, and single-link operation at about 30 Mbps, so STR doubles the throughput that EMLSR provides.
  • STR keeps queuing, access, and end-to-end delays lower than EMLSR in every scenario the paper reports, with the gap largest under high offered load.
  • Raising the MCS or channel bandwidth increases saturated throughput in both modes, but STR stays ahead at every setting, and the load at which the network saturates is consistently higher for STR.
  • Adding interfering single-link devices reduces throughput by similar amounts in both modes, so the STR advantage persists under the tested interference configurations.
  • For applications that need peak throughput or low latency, the paper's results favor STR; EMLSR remains a candidate only where power consumption is the dominant constraint, an aspect the paper does not quantify.

Reading between the lines

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

  • I infer from the simulation pattern that an adaptive policy that switches between STR and EMLSR based on queue load or interference could beat either fixed mode; the paper hints at such dynamic allocation in its future-work section but does not simulate it.
  • The roughly equal throughput drop when a single interferer is added to either link suggests the simulator treats both links as near-identical; in a real deployment with asymmetric channel quality, the choice between STR and EMLSR might depend more on link conditions than on mode alone.
  • The energy-efficiency advantage for EMLSR is not yet supported by the paper's own experiments; adding an energy model would be a direct way to test when EMLSR's lower throughput is worth its lower power.
  • I infer that the STR advantage is likely largest in saturated, high-load networks; at very light load the two modes may be nearly indistinguishable, which is not highlighted in the paper's figures.
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 compares three Wi-Fi 7 modes—STR, EMLSR, and SLO—using ns-3.41 simulations in five scenario families: a base network, varying network size, varying MCS, varying bandwidth, and interference. The authors report saturated throughput, queuing delay, access delay, and end-to-end delay, and conclude that STR consistently outperforms EMLSR. The paper also claims in the abstract that EMLSR balances energy efficiency with responsiveness, but no energy metrics are collected.

Significance. If the results hold, the paper provides a useful comparative data point for mode selection in Wi-Fi 7 MLO, namely that STR dominates EMLSR in saturated throughput and queuing/end-to-end delay across the tested parameter ranges. The availability of the simulation scripts in a public repository is a genuine reproducibility strength, and the use of an open-source simulator with no fitted parameters avoids circularity. However, the significance is limited by the absence of statistical uncertainty quantification, the lack of validation of the ns-3 EMLSR/STR model, and the overly broad conclusion that ignores the paper's own access-delay exceptions.

major comments (4)
  1. [IV and Section III-C/D] The statement in Section IV that 'STR consistently outperformed EMLSR' is contradicted by the paper's own access-delay results. Section III-C states that, from Fig. 10, STR should be favored for access delay only when mcs={2,4}, the modes are equal at mcs=6, and EMLSR should be favored at mcs=8. Section III-D states, from Fig. 14, that EMLSR should be favored for access delay when channelWidth={40,80}. Since access delay is one of the paper's primary latency metrics, the conclusion must be qualified: STR outperforms EMLSR in throughput, queuing delay, and end-to-end delay, but access-delay superiority is regime-dependent.
  2. [Section III (all scenarios)] No statistical uncertainty is reported. The paper does not state the number of simulation runs, the random seeds, or any confidence intervals, and all comparisons are based on point estimates of mean throughput and delay. In stochastic discrete-event simulations such as ns-3, run-to-run variability can be substantial; without this information, it is unclear whether the reported throughput differences (e.g., 80 vs. 20 Mbps at mcs=2 in Fig. 8) are statistically meaningful. The authors should add repeated runs with different seeds and report means with confidence intervals or at least state the number of runs and the seed policy.
  3. [Abstract and Section IV] The energy-efficiency claim in the abstract is unsupported. The abstract states that 'EMLSR balances energy efficiency with responsiveness,' but no energy consumption metric, model, or measurement appears anywhere in the paper. In fact, Section IV lists 'implementing energy efficiency comparisons' as future work. The energy-efficiency statement should be removed from the abstract and introduction, or the paper should add actual energy-efficiency simulation results.
  4. [Section II and Section III] The simulation model is not validated. The paper relies on ns-3.41's native EMLSR and EHT support, but provides no comparison with measurements, analytical bounds, or an independent reference model. Because the main ranking STR > EMLSR depends on the fidelity of the single-radio switching constraint and dual-link operation, a simulator artifact could change the ranking. At minimum, the paper should discuss known limitations of the ns-3.41 MLO implementation, cite its validation literature if any applies to this configuration, and provide a sanity check such as a comparison with saturation-throughput bounds.
minor comments (5)
  1. [Section III-B] The text repeatedly uses 'throughout' where 'throughput' is intended (e.g., in the descriptions of Figs. 5-7). Additionally, the unit of lambda (offered load) is never defined; please state whether it is packets per second, packets per slot, or another quantity.
  2. [Section III-A] The phrase 'payload of 1500 packets' should read 'payload of 1500-byte packets'; as written, it suggests a packet count rather than a payload size.
  3. [Section III-C] The mapping of MCS values to modulations appears non-standard: mcs=8 is described as 128-QAM, whereas common 802.11 tables list mcs=8 as 256-QAM (or a different rate depending on the PHY). Please verify the MCS table used for the simulations and correct the text if needed.
  4. [Section III-E] Scenarios 3 and 4 are single-mode experiments (STR-only and EMLSR-only) and do not provide a direct STR-vs-EMLSR comparison. They should be relabeled or the text should explain how they support the comparative conclusion. Also, 'there is to contention for access' appears to be a typographical error.
  5. [Section III-D] The sentence 'EMLSR should be favored this time when channelWidth={40,80}' is missing a closing phrase such as 'for applications requiring lower access delay'; although the preceding sentence supplies the context, the abrupt ending hurts readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper reports observed ns-3 simulation outputs with no fitted parameters and no load-bearing self-citation; the internal contradiction about access delay is a consistency issue, not circularity.

full rationale

The paper is a simulation study comparing STR, EMLSR, and SLO in ns-3.41. No model parameter is fitted to the data it then 'predicts'; the reported throughput and delay curves are direct outputs of the simulator under the stated configurations. The only self-referential item is the project's own GitHub repository [6], which is used for code availability and is not load-bearing for any conclusion. The external references [1]-[4], [7], [8], [9], [10]-[12] are prior work or tool citations, not self-citations by the present authors, and none is invoked to force the STR-versus-EMLSR ranking. The central claim that 'STR consistently outperformed EMLSR' is an empirical summary of the simulations, not an assumption embedded in the inputs. A separate concern is that the paper's own Sections III-C and III-D state that EMLSR should be favored for access delay when mcs=8 and when channelWidth={40,80}, which contradicts the unqualified 'consistently' in the conclusion; however, that is an internal-consistency or correctness risk, not a circularity. Similarly, the lack of validation of the ns-3 EMLSR/STR implementation against measurements is a validity threat, not a circular derivation. No equation or definition reduces the claimed result to its own inputs, and no fitted parameter is renamed as a prediction. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no fitted parameters or invented entities. Its central claim rests on the fidelity of the ns-3 simulator and on the representatives of the scenario parameters, both unquestioned in the text.

assumptions (3)
  • domain assumption ns-3 version 3.41 correctly models Wi-Fi 7 EHT features including STR and EMLSR.
    Section II states that the release introduced EHT features and native EMLSR support; the paper does not validate the model against measurements or a reference implementation.
  • domain assumption The chosen simulation scenarios (MCS, BW, interference levels, offered load ranges) are representative of real Wi-Fi 7 deployments.
    Section III describes scenarios but provides no justification for the parameter ranges or a link to real traffic patterns.
  • domain assumption Simulation outputs are deterministic or sufficiently stable without multiple seeds.
    No seeds, number of runs, or confidence intervals are reported, yet throughput and delay numbers are quoted to the nearest integer.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Comparison of STR and EMLSR Performance in Wi-Fi 7 MLO." pith.science (2026). https://pith.science/paper/UEJ4BKYT

@misc{pith2026250104149,
  author       = {Pith},
  title        = {Pith review of: Comparison of STR and EMLSR Performance in Wi-Fi 7 MLO},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UEJ4BKYT}},
  note         = {Machine review of arXiv:2501.04149}
}
read the original abstract

This project compares the performance of simultaneous transmit and receive (STR) and enhanced multi-link single radio (EMLSR) within Multi-Link Operation (MLO) in Wi-Fi 7 networks. Using the ns-3 simulator, we evaluate both techniques under various scenarios, including changes in modulation coding scheme (MCS), bandwidth, link quality, and interference levels. Key performance metrics such as latency, throughput, and energy efficiency are analyzed to determine the trade-offs between STR and EMLSR. The results demonstrate that STR achieves higher throughput and lower latency due to dual-link utilization, making it suitable for high-load environments. In contrast, EMLSR balances energy efficiency with responsiveness, making it advantageous for power-sensitive applications. This analysis provides insights into the strengths and limitations of STR and EMLSR, guiding optimal deployment strategies for future Wi-Fi 7 networks.

Figures

Figures reproduced from arXiv: 2501.04149 by the authors.

Figure 1
Figure 1. Throughput Comparison for SLO, STR, and EMLSR [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Queuing Delay Comparison for SLO, STR, and [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Access Delay Comparison for SLO, STR, and [PITH_FULL_IMAGE:figures/full_fig_p002_3.png] view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: End to End Delay Comparison for SLO, STR, and [PITH_FULL_IMAGE:figures/full_fig_p002_4.png]
Figure 8
Figure 8. Figure 8: STR vs EMLSR: Throughput vs Offered Load with [PITH_FULL_IMAGE:figures/full_fig_p003_8.png]
Figure 9
Figure 9. Figure 9: STR vs EMLSR: Mean Queuing Delay vs Offered [PITH_FULL_IMAGE:figures/full_fig_p003_9.png]
Figure 10
Figure 10. Figure 10: STR vs EMLSR: Mean Access Delay vs Offered [PITH_FULL_IMAGE:figures/full_fig_p003_10.png]
Figure 11
Figure 11. Figure 11: STR vs EMLSR: Mean E2E Delay vs Offered Load [PITH_FULL_IMAGE:figures/full_fig_p003_11.png]
Figure 12
Figure 12. Figure 12: STR vs EMLSR: Throughput vs Offered Load with [PITH_FULL_IMAGE:figures/full_fig_p004_12.png]
Figure 15
Figure 15. Figure 15: STR vs EMLSR: Mean E2E Delay vs Offered Load [PITH_FULL_IMAGE:figures/full_fig_p004_15.png]
Figure 16
Figure 16. Figure 16: Multi-MLD STR Network: Throughput vs Offered [PITH_FULL_IMAGE:figures/full_fig_p005_16.png]
Figure 19
Figure 19. Figure 19: Multi-MLD EMLSR Network: Throughput vs Offered Load under Symmetric Interference 2) Scenario 2 - Multi-MLD Network under Symmetric In￾terference: The second scenario again considers a network consisting of one AP with five MLDs, all operating under either STR or EMLSR…
Figure 23
Figure 23. Figure 23: Single-MLD STR Network: End to End Delay [PITH_FULL_IMAGE:figures/full_fig_p006_23.png]
Figure 26
Figure 26. Figure 26: Single-MLD eMLSR Network: Access Comparison [PITH_FULL_IMAGE:figures/full_fig_p006_26.png]
Figure 27
Figure 27. Figure 27: Single-MLD eMLSR Network: End to End Comparison technique performs under varying conditions, such as changes in modulation and coding schemes (MCS), channel bandwidth, and interference levels. Our results exhibiting the throughput improvements and latency reductions o…
Figure 24
Figure 24. Figure 24: shows the throughput for all the stations tested, and as our nSTA’s increase the saturation of throughput decreases. For example, at nSTa = 20 the throughput saturates at a lower value of lambda. The next plots [PITH_FULL_IMAGE:figures/full_fig_p006_24.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 12 canonical work pages

  1. [1]

    Delay in multi-link operation in ns-3: Validation and impact of traffic splitting,

    M. Shen, J. Zhang, H. Yin, S. Roy, and Y . Gao, “Delay in multi-link operation in ns-3: Validation and impact of traffic splitting,” in Proceedings of the 2024 Workshop on Ns-3 , ser. WNS3 ’24. New York, NY , USA: Association for Computing Machinery, 2024, p. 19–26. [Online]. Available: https://doi.org/10.1145/3659111.3659116

    arXiv 2024

  2. [2]

    Model-free dynamic traffic steering for multi-link operation in ieee 802.11be,

    C.-L. Tai, M. Eisen, D. Akhmetov, D. Das, D. Cavalcanti, and R. Sivaku- mar, “Model-free dynamic traffic steering for multi-link operation in ieee 802.11be,” in 2024 IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN) , 2024, pp. 44–49

    work page 2024

  3. [3]

    Ieee 802.11be network throughput optimization with multilink operation and ap controller,

    L. Zhang, H. Yin, S. Roy, L. Cao, X. Gao, and V . Sathya, “Ieee 802.11be network throughput optimization with multilink operation and ap controller,” IEEE Internet of Things Journal , vol. 11, no. 13, pp. 23 850–23 861, 2024

    work page 2024

  4. [4]

    Enhanced multilink single-radio operation for the next-generation ieee 802.11 be wi-fi systems,

    X. Lan, X. Zu, and J. Yang, “Enhanced multilink single-radio operation for the next-generation ieee 802.11 be wi-fi systems,” Security and Communication Networks , vol. 2022, no. 1, p. 7018360,

    work page 2022

  5. [5]

    nsnam. ns-3.41. [Online]. Available: https://www.nsnam.org/releases/ ns-3-41/

    work page

  6. [6]

    G. 5. ee595-f24-group5. [Online]. Available: https://github.com/ kevinho-uw/ee595-f24-group5

    work page

  7. [7]

    Performance of 802.11be wi-fi 7 with multi-link operation on ar applications,

    M. Alsakati, C. Pettersson, S. Max, V . N. Moothedath, and J. Gross, “Performance of 802.11be wi-fi 7 with multi-link operation on ar applications,” 2023. [Online]. Available: https://arxiv.org/abs/2304. 01693

    work page 2023

  8. [8]

    Wi-fi multi-link operation: An experimental study of latency and throughput,

    M. Carrascosa-Zamacois, G. Geraci, E. Knightly, and B. Bellalta, “Wi-fi multi-link operation: An experimental study of latency and throughput,” IEEE/ACM Transactions on Networking , vol. 32, no. 1, pp. 308–322, 2024

    work page 2024

Show all 13 references
  1. [9]

    Networks

    S. Networks. Mcs index, modulation and coding scheme. [Online]. Available: https://mcsindex.net/

  2. [10]

    Ns3-ai: Fostering artificial intelligence algorithms for networking research,

    H. Yin, P. Liu, K. Liu, L. Cao, L. Zhang, Y . Gao, and X. Hei, “Ns3-ai: Fostering artificial intelligence algorithms for networking research,” in Proceedings of the 2020 Workshop on Ns-3 , ser. WNS3 2020. New York, NY , USA: Association for Computing Machinery, 2020, p. 57–64....

    2020

  3. [11]

    A federated reinforcement learning framework for link activation in multi-link wi-fi networks,

    R. Ali and B. Bellalta, “A federated reinforcement learning framework for link activation in multi-link wi-fi networks,” 2023. [Online]. Available: https://arxiv.org/abs/2304.14720

    2023 arXiv

  4. [12]

    Channel selection for wi-fi 7 multi-link operation via optimistic-weighted vdn and parallel transfer reinforcement learning,

    P. E. Iturria-Rivera, M. Chenier, B. Herscovici, B. Kantarci, and M. Erol-Kantarci, “Channel selection for wi-fi 7 multi-link operation via optimistic-weighted vdn and parallel transfer reinforcement learning,” 2023. [Online]. Available: https://arxiv.org/abs/2307.05419

    2023 arXiv

  5. [2022]

    Available: https://onlinelibrary.wiley.com/doi/abs/10

    [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10. 1155/2022/7018360

    2022

Pith tools

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