Pith. sign in

REVIEW 3 major objections 4 minor 14 references

Relay Selection and User Equipment Admission in Resource-Efficient NextG Sidelink Communications

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read In sidelink networks, dynamic per-user weights turn a greedy relay-selection rule into a fair one: more outer users are admitted, more evenly over time, and total data rate stays close to the rate-maximizing greedy schedule.

desk verdict A clean, plausible greedy plus weight-based fair scheduling study for sidelink relay selection, but the fairness claim rests on a single simulation setup and a truncated exhaustive-search baseline. read the letter →

arxiv 2506.03328 v1 pith:FZSXEFIC submitted 2025-06-03 cs.NI cs.DCcs.ITeess.SPmath.IT

Yalin E. Sagduyu , Tugba Erpek , Sastry Kompella , Kemal Davaslioglu This is my paper
classification cs.NIcs.DCcs.ITeess.SPmath.IT
keywords sidelinkcommunications5G/6GrelayselectionUEadmissiongreedyalgorithmfairschedulingweightedsumratenetworkoptimization
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 tackles the relay-association problem in 5G/6G sidelink networks, where outer user devices that cannot reach a base station send traffic through inner relay devices. It formulates the choice of which outer user attaches to which inner relay as a non-convex weighted-sum-rate optimization under channel, interference, and traffic constraints. Against an approximate exhaustive-search upper bound, the paper shows that a greedy link-selection algorithm sustains high sum rates and scales to many users, but leaves some outer users unserved. Its central claim is that modifying the greedy objective with per-user weights that grow while a user waits, or track queue length, raises the mean admission ratio of outer users and lowers its variance across users while keeping the achievable sum rate close to the greedy level. A distributed implementation using discovery messages and channel-state reports completes the picture.

What carries the argument

The carrying mechanism is the per-slot weight-update rule for each outer UE, used inside the weighted sum-rate objective. For the wait-time version, $w_i \leftarrow w_i + 1$ when UE $i$ is not activated and $w_i \leftarrow \max(w_i - 1, 1)$ when it is; for the queue version, $w_i \leftarrow w_i + \lambda_i$ when not activated and $w_i \leftarrow \max(w_i + \lambda_i - r_i, 0)$ when it is. These weights are what the greedy link selector maximizes, so the fairness mechanism is not an extra constraint but a time-varying reweighting of the original objective. The admission ratio of an outer UE is the fraction of slots in which it is selected, and the paper evaluates fairness as the mean and variance of this ratio over time and across UEs.

What would settle it

In the paper's 500-slot simulation, make one outer UE's channels 10 dB weaker to every inner UE while keeping them positive; the wait-time fair rule predicts that UE's admission ratio rises above greedy's as its weight accumulates, and if it never rises the fairness claim reduces to serving UEs greedy would already admit.

Watch

Extended reading notes

Core claim

The paper's core discovery is that fairness in sidelink relay admission does not require a different scheduler; it can be obtained by feeding dynamic weights into the same greedy rule. In the wait-time variant, an outer UE that is not activated in a slot has its weight incremented, while an activated UE's weight is reduced but floored at one. In the queue variant, weights evolve as queue backlogs, increasing by arrivals and decreasing by service. When the greedy algorithm maximizes the weighted sum rate with these weights, outer UEs that would otherwise be starved are eventually selected, and the numerical results show the mean admission ratio rises and the variance across UEs falls relative to unweighted greedy, with greedy retaining a modest edge in achievable sum rate. The paper also reports that greedy selection outperforms the truncated exhaustive search once the number of outer UEs exceeds eight, and that the fair algorithms also keep maximum wait time bounded under the wait-time rule.

Load-bearing premise

The load-bearing premise is that 50,000 sampled schedules approximate the global optimum closely enough to serve as the rate upper bound, even though the full search space is much larger once $n_o$ exceeds five.

Editorial extensions

If this is right

  • Under the fair weight-update rules, the mean admission ratio of outer UEs increases and its variance decreases, so the network serves a larger and more even share of connection requests over time.
  • The sum rate achievable by the fair algorithms stays close to the pure greedy algorithm, with the queue-based rule the closest, so fairness is obtained without a large throughput penalty in the simulated regimes.
  • The greedy algorithm's achievable sum rate grows monotonically with the number of outer UEs up to at least 250, meaning the relay-selection method scales beyond the sizes where exhaustive search is feasible.
  • The distributed procedure can be run with about $4 n_o + 1$ message exchanges before collision-resolution retries, using discovery messages, CSI measurement, and gNodeB broadcast of assignments.
  • The weight-rule idea is positioned to extend via backpressure-style differential backlogs to routing over more than two hops.

Reading between the lines

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

  • A testable extension not pursued in the paper: the same dynamic-weight mechanism could be applied to other constrained assignment problems where a greedy rate maximizer starves low-priority entities, and a formal convergence or throughput-optimality analysis would strengthen the simulation-based evidence.
  • Because the wait-time rule keeps maximum wait time low while the queue-based rule better matches traffic, an adaptive rule that switches between the two could reduce the queue-based rule's growing maximum wait time without sacrificing its delay performance.
  • The fairness result is shown for equal numbers of inner and outer UEs in a two-ring topology; an honest stress test is an asymmetric placement where one outer UE is distant from all relays, where the model predicts delayed but eventual admission whenever any positive-rate link exists.
  • The reported greedy-beats-exhaustive crossover for $n_o > 8$ is sensitive to the 50,000-schedule cap; with an exact solver for small $n_o$ the crossover point could move, and the fair-vs-greedy rate gap might widen or narrow accordingly.
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.

Referee Report

3 major / 4 minor

Summary. The paper studies relay selection and outer-UE admission in two-hop 5G/6G sidelink networks. It formulates the assignment of outer UEs to inner relay UEs as a weighted sum-rate maximization with per-hop capacity, interference, and traffic constraints, notes that the problem is non-convex and that exhaustive enumeration scales as (n_o+1)^{n_o}, and then proposes a greedy link-addition algorithm and two fairness-oriented variants whose weights are updated from waiting time or queue length. Simulation results are used to argue that the greedy algorithm maintains high sum rate while the fair algorithms increase the mean admission ratio and reduce its variance across UEs at modest sum-rate cost. The paper also sketches a distributed CSI/discovery procedure with a message-count bound.

Significance. If the results are robust, the paper addresses a practically relevant trade-off in sidelink relay selection: maximizing throughput while avoiding starvation of outer UEs. The paper is transparent about the exponential complexity of the global search and about the limits of its exhaustive baseline, and the simulation setup is described in enough detail to be reproduced. However, the central quantitative claims rest on two fragile pillars: the 'exhaustive search' benchmark is not exhaustive for n_o > 5, and the fairness improvement is demonstrated at a single parameter point with no sensitivity analysis or statistical uncertainty reporting. The work is a useful algorithmic study, but as it stands the evidence for the load-bearing claims is incomplete.

major comments (3)
  1. [Sec. III, Figs. 4-7] The 'exhaustive search' baseline is not the true global optimum for n_o > 5 because the search space (n_o+1)^{n_o} exceeds the 50,000 schedules searched, as the text itself acknowledges. The saturation of the exhaustive-search curve starting at n_o = 6 and the claim in Sec. IV (Fig. 6) that the greedy algorithm achieves higher rates than exhaustive search for n_o > 8 are therefore conclusions about a truncated heuristic baseline, not about optimality. Please either replace this baseline with a valid optimal method for at least n_o up to 6 (e.g., branch-and-bound or complete enumeration with pruning) or relabel the baseline as a 'limited-search heuristic' and remove language that equates it with the global upper bound.
  2. [Sec. V, Figs. 9-11] The central fairness claim is supported by a single simulation configuration: 500 slots, 1000 repetitions, equal numbers of inner and outer UEs, and weight updates of +1/-1 (wait time) or +lambda_i and -r_i (queue). No horizon, step-size, or traffic-intensity sensitivity is reported. The wait-time step is in the same units as the objective weights while the rates r_i are in bit/s/Hz and can be much smaller than 1, so a short starvation episode can make a UE's weight dominate the weighted sum rate and trivially force admission; conversely, 500 slots may be too short for the queue-based rule to lift a persistently starved UE under heavier traffic. Please add sensitivity sweeps over the number of slots, the weight-update step size, arrival rates, and asymmetry between inner and outer UE counts, and show whether the reported mean/variance/rate trade-off persists.
  3. [Sec. V, Figs. 9-11] No error bars, confidence intervals, or other uncertainty measures are reported, even though the variance of the admission ratio is itself a second-order statistic estimated from 1000 repetitions and is therefore noisy. The visible differences between the greedy and fair curves, particularly at small n_o, could be within Monte Carlo error. Please report standard errors or confidence bands, and if the same channel realizations and traffic draws are used for all algorithms, also state whether paired comparisons were employed.
minor comments (4)
  1. [Throughout] There are several typographical and spacing errors, including 'of5G networks', 'an Unparalleled', 'problem of determining' in the abstract, 'n >5' where 'n_o >5' is meant in Sec. III, and 'determines with UE links' where 'determines which UE links' is meant in Sec. VI.
  2. [Fig. 5] The legend for the curve 'relay traffic rate = 0.5, path loss = 2' omits the 'Exhaustive Search' prefix used by the other curves, making the figure harder to read.
  3. [Sec. V, Fig. 12] The greedy algorithm is absent from the maximum-wait-time plot because UEs that are never activated have undefined wait times; the text should state this exclusion explicitly, since it affects how the fair algorithms' advantage is interpreted.
  4. [Sec. VI] The message-count derivation should clarify whether the n_o messages in step 3 are counted per inner UE or per outer UE, and the sentence 'If there are n_o outer UEs and n_o inner UEs)' contains an extra parenthesis that should be removed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's performance claims are simulation-based and do not reduce to their inputs by construction.

full rationale

The paper's load-bearing claims are (i) greedy achieves high sum rate at the cost of fairness, and (ii) wait-time/queue-based weight updates improve the mean and variance of admission ratio with modest sum-rate loss. Both claims are established by Monte Carlo simulation over 500 slots and 1000 repetitions, not by deriving a target result from fitted parameters. The fair weight updates (w_i <- w_i + 1 / max(w_i - 1, 1) and w_i <- w_i + lambda_i / max(w_i + lambda_i - r_i, 0)) are explicit algorithm definitions; the measured admission-ratio improvement is an empirical consequence of those rules, not a quantity used to define the rules. The phrase "greedy algorithm can sustain higher achievable sum rate as it is designed to do" acknowledges a design sanity check rather than presenting a derived prediction. The exhaustive-search baseline is explicitly acknowledged to be incomplete for n_o > 5 because (n_o+1)^n_o exceeds the 50,000 searched schedules; this is a validity limitation, not a circular step. The only self-citation, [13], supports the background claim that single-hop rate optimization is nonconvex and is not load-bearing for the paper's fairness or sum-rate conclusions. Since no prediction is algebraically identical to an input and no fitted parameter is renamed as a result, there is no circularity.

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

No entities are invented. The central claims rest on standard system-model assumptions and on simulation parameters chosen by hand. The most fragile assumption is the truncated exhaustive-search baseline, which the paper itself flags.

free parameters (4)
  • path loss coefficient alpha = 2 and 4
    Chosen as simulation settings; the paper uses these values to show the effect of signal and interference attenuation (Sec. III).
  • relay traffic rate r_j^(2) = drawn uniformly from [0, r_max] with r_max = 0, 0.1, 0.5
    Simulation parameter representing inner UE background traffic; the results in Figs. 5 and 7 depend on it.
  • simulation slots and repetitions = 500 slots, 1000 averaging runs
    Used to compute admission-ratio mean and variance in Sec. V; differences are not accompanied by confidence intervals.
  • arrival rate for queue-based algorithm = mean total arrival rate 0.5, Bernoulli process
    Defines the queueing load for the queueing-delay results in Fig. 13.
assumptions (4)
  • domain assumption PC5 and Uu interfaces use different frequency bands, so Hop 1 and Hop 2 do not interfere.
    Stated in Sec. II; removes cross-hop interference and simplifies the sum-rate objective.
  • domain assumption The gNodeB has multiple antennas and serves multiple Hop 2 UEs without interference, while each Hop 2 UE has one interface and can relay for at most one Hop 1 UE.
    Sec. II constraints C2 and C3 define the feasible assignment set.
  • domain assumption Link capacities follow the Shannon formula with SINR using only path-loss attenuation; small-scale fading is not specified.
    Sec. II equations C4 and C5; the numerical results depend on this channel model.
  • ad hoc to paper A 50,000-schedule exhaustive search is treated as the global optimum for sum rate.
    Sec. III; the paper admits the search space exceeds 50,000 for n_o > 5, so the baseline is approximate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Relay Selection and User Equipment Admission in Resource-Efficient NextG Sidelink Communications." pith.science (2026). https://pith.science/paper/FZSXEFIC

@misc{pith2026250603328,
  author       = {Pith},
  title        = {Pith review of: Relay Selection and User Equipment Admission in Resource-Efficient NextG Sidelink Communications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FZSXEFIC}},
  note         = {Machine review of arXiv:2506.03328}
}
read the original abstract

5G/6G sidelink communications addresses the challenge of connecting outer UEs, which are unable to directly access a base station (gNodeB), through inner UEs that act as relays to connect to the gNodeB. The key performance indicators include the achievable rates, the number of outer UEs that can connect to a gNodeB, and the latency experienced by outer UEs in establishing connections. We consider problem of determining the assignment of outer UEs to inner UEs based on the channel, interference, and traffic characteristics. We formulate an optimization problem to maximize a weighted sum rate of UEs, where weights can represent priority, waiting time, and queue length. This optimization accommodates constraints related to channel and interference characteristics that influence the rates at which links can successfully carry assigned traffic. While an exhaustive search can establish an upper bound on achievable rates by this non-convex optimization problem, it becomes impractical for larger number of outer UEs due to scalability issues related to high computational complexity. To address this, we present a greedy algorithm that incrementally selects links to maximize the sum rate, considering already activated links. This algorithm, although effective in achieving high sum rates, may inadvertently overlook some UEs, raising concerns about fairness. To mitigate this, we introduce a fairness-oriented algorithm that adjusts weights based on waiting time or queue length, ensuring that UEs with initially favorable conditions do not unduly disadvantage others over time. We show that this strategy not only improves the average admission ratio of UEs but also ensures a more equitable distribution of service among them, thereby providing a balanced and fair solution to sidelink communications.

Figures

Figures reproduced from arXiv: 2506.03328 by the authors.

Figure 1
Figure 1. Sidelink communications. Sidelink communications is highly important for next￾generation communication systems for several reasons: (i) Enhanced connectivity: Sidelink allows devices to communi￾cate directly without relying on a centralized network infras￾tructure, which is crucial in environments where traditional communication networks are unavailable or unreliable. (ii) Re￾silience and reliability: Sidelink can m… view at source ↗
Figure 2
Figure 2. Relay selection for sidelink communications. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. Sum rate achieved by exhaustive search, random [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figures from the paper (10 more)
Figure 5
Figure 5. Figure 5: Sum rate achieved by exhaustive search for different [PITH_FULL_IMAGE:figures/full_fig_p003_5.png]
Figure 3
Figure 3. Figure 3: Network topology. The resulting achievable sum rate (measured in bit/s/Hz) is shown in [PITH_FULL_IMAGE:figures/full_fig_p003_3.png]
Figure 6
Figure 6. Figure 6: Sum rate achieved by greedy algorithm. 2 3 4 5 6 7 8 9 10 Number of Outer UEs 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Achievable Sum Rate Greedy Algorithm, relay traffic rate = 0.1, path loss = 2 Exhaustive Search, relay traffic rate = 0.1, path los…
Figure 7
Figure 7. Figure 7: Sum rate achieved by greedy algorithm for different [PITH_FULL_IMAGE:figures/full_fig_p004_7.png]
Figure 8
Figure 8. Figure 8: Scaling of sum rate achieved by greedy algorithm with [PITH_FULL_IMAGE:figures/full_fig_p004_8.png]
Figure 9
Figure 9. Figure 9: Mean of admission ratio achieved by wait time-based [PITH_FULL_IMAGE:figures/full_fig_p005_9.png]
Figure 10
Figure 10. Figure 10: Variance of admission ratio achieved by wait time [PITH_FULL_IMAGE:figures/full_fig_p005_10.png]
Figure 11
Figure 11. Figure 11: Sum rate achieved by wait time-based and queue [PITH_FULL_IMAGE:figures/full_fig_p005_11.png]
Figure 12
Figure 12. Figure 12: Maximum wait time incurred by wait time-based and [PITH_FULL_IMAGE:figures/full_fig_p006_12.png]
Figure 13
Figure 13. Figure 13: Average queueing delay incurred by queue-based fair [PITH_FULL_IMAGE:figures/full_fig_p006_13.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

  1. [13]

    Distributed stochastic power control in ad hoc networks: a nonconvex optimization case,

    L. Yang, Y . E. Sagduyu, J. Zhang, and J. H. Li, “Distributed stochastic power control in ad hoc networks: a nonconvex optimization case,” EURASIP Journal on Wireless Communications and Networking, 2012

    work page 2012

  2. [1]

    Who needs basestations when we have sidelinks?

    V . Weerackody, K. Benson, and S. Roy, “Who needs basestations when we have sidelinks?”Global Communications, 2023

    work page 2023

  3. [2]

    V2X in 3GPP standardization: NR sidelink in release-16 and beyond,

    M. Harounabadi, D. M. Soleymani, S. Bhadauria, M. Leyh, and E. Roth- Mandutz, “V2X in 3GPP standardization: NR sidelink in release-16 and beyond,”IEEE Communications Standards Magazine, 2021

    work page 2021

  4. [3]

    3GPP NR sidelink transmissions toward 5G V2X,

    S.-Y . Lien, D.-J. Deng, C.-C. Lin, H.-L. Tsai, T. Chen, C. Guo, and S.- M. Cheng, “3GPP NR sidelink transmissions toward 5G V2X,”IEEE Access, 2020

    work page 2020

  5. [4]

    5G advanced: Sidelink evolution,

    K. Ganesan, “5G advanced: Sidelink evolution,”IEEE Communications Standards Magazine, 2023

    work page 2023

  6. [5]

    Sidelink evolution toward 5G-A/6G future considerations for standardization of group communi- cations,

    R. Shrivastava, S. Hegde, and O. Blume, “Sidelink evolution toward 5G-A/6G future considerations for standardization of group communi- cations,”IEEE Communications Standards Magazine, 2023

    work page 2023

  7. [6]

    On the design of sidelink for cellular V2X: A literature review and outlook for future,

    A. Bazzi, A. O. Berthet, C. Campolo, B. M. Masini, A. Molinaro, and A. Zanella, “On the design of sidelink for cellular V2X: A literature review and outlook for future,”IEEE Access, 2021

    work page 2021

  8. [7]

    Open source-based over-the-air 5G new radio sidelink testbed,

    M. Elkadi, D. Kim, E. Ahmed, A. Le, M. Sadeghi, P. Russell, and B. Ryu, “Open source-based over-the-air 5G new radio sidelink testbed,” inIEEE Military Communications Conference (MILCOM), 2023

    work page 2023

Show all 14 references
  1. [8]

    System- level evaluation of 5G NR UE-based relays,

    S. Gamboa, A. B. Mosbah, W. Garey, C. Liu, and R. Rouil, “System- level evaluation of 5G NR UE-based relays,” inIEEE MILCOM, 2023

    2023

  2. [9]

    Sideseeker: Contention-based distributed relay finding for sidelink mesh networks,

    T. Wang, X. Wang, and Y .-B. Lin, “Sideseeker: Contention-based distributed relay finding for sidelink mesh networks,”IEEE Wireless Communications Letters, 2024

    2024

  3. [10]

    Generalized multi-hop NR sidelink relay for future V2X communication,

    Q. Fu and J. Liu, “Generalized multi-hop NR sidelink relay for future V2X communication,”IEEE/ACM Transactions on Networking, 2023

    2023

  4. [11]

    Performance characterization of joint communication and sens- ing with beyond 5G NR-V2X sidelink,

    N. Decarli, S. Bartoletti, A. Bazzi, R. A. Stirling-Gallacher, and B. M. Masini, “Performance characterization of joint communication and sens- ing with beyond 5G NR-V2X sidelink,”IEEE Transactions on Vehicular Technology, 2024

    2024

  5. [12]

    NR sidelink mode 2 in unlicensed bands: Throughput model & validation,

    V . Weerackody, H. Yin, and S. Roy, “NR sidelink mode 2 in unlicensed bands: Throughput model & validation,”IEEE Transactions on Com- munications, 2024

    2024

  6. [14]

    Stability properties of constrained queueing systems and scheduling policies for maximum throughput in multihop radio networks,

    L. Tassiulas and A. Ephremides, “Stability properties of constrained queueing systems and scheduling policies for maximum throughput in multihop radio networks,”IEEE Transactions on Automatic Control, 1992

    1992

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.