Pith. sign in

REVIEW 4 major objections 5 minor 46 references

Uplink Rate-Splitting Multiple Access for Mobile Edge Computing with Short-Packet Communications

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

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

Pith's one-line read Uplink RSMA with one split stream achieves higher successful computation probability and lower latency than NOMA in short-packet mobile edge computing.

desk verdict A legitimate new combination of uplink RSMA + FBL + MEC, but the convergence proof has a sign error and the surrogates are unverified, so the reported gains are plausible yet not established. read the letter →

arxiv 2502.04827 v1 pith:2NXR6Y3V submitted 2025-02-07 cs.IT eess.SPmath.IT

Jiawei Xu , Yumeng Zhang , Yunnuo Xu , Bruno Clerckx This is my paper · ORCID
classification cs.ITeess.SPmath.IT
keywords rate-splittingmultipleaccessmobileedgecomputingfiniteblocklengthshort-packetcommunicationssuccessfulcomputationprobabilityNOMAcomparisontaskoffloadingpowerallocation
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 argues that uplink rate-splitting multiple access (RSMA) is a better multiple-access scheme than non-orthogonal multiple access (NOMA) for mobile edge computing (MEC) when task offloading uses short packets with finite blocklengths. It builds a two-user MEC model in which one user splits its task into two streams transmitted alongside the other user's unsplit stream, derives the Successful Computation Probability (SCP) under finite-blocklength constraints, and shows numerically that RSMA attains a higher SCP than NOMA and reaches the same SCP with a shorter blocklength, which translates to lower latency. The optimization jointly chooses how much of each task to offload, how to split one user's offloaded task, and how to allocate transmit power between the RSMA streams, using alternating optimization with successive convex approximation. If the claim holds, RSMA becomes a practical candidate for reliable low-latency edge offloading in future wireless systems.

What carries the argument

The carrier of the argument is the two-user uplink RSMA structure in which user 1's task is split into two independently encoded streams, $s_{1,1}$ and $s_{1,2}$, while user 2 remains unsplit, decoded at the edge server in the order $s_{1,1} \rightarrow s_2 \rightarrow s_{1,2}$. This decoding order gives the first stream a manageable signal-to-interference-plus-noise ratio, lets the second user be decoded after partial interference cancellation, and finally decodes the interference-free remainder of user 1; the paper relies on a prior finite-blocklength analysis showing that this order yields much lower error probability than NOMA's fixed order. On top of this, the finite-blocklength achievable-rate expression $r \approx C(\gamma) - \sqrt{V(\gamma)/N}\,Q^{-1}(\epsilon)$ turns each stream's error probability into an explicit function of SINR, data size, and blocklength $N$, which makes the SCP an optimizable quantity. The optimization then uses the small-error approximation $1 - (2\varepsilon_a + 2\varepsilon_b + \varepsilon_c)$ and the Chernoff bound $\exp(-x^2/2) \ge Q(x)$ to turn SCP maximization into minimizing a sum of exponentials, solved by alternating optimization with successive convex approximation using first-order Taylor surrogates for the non-convex SINR and data-size constraints.

What would settle it

Re-derive the SCP curves for the same channel realizations using the exact joint error probability $(1-\varepsilon_a)(1-\varepsilon_b)(1-\varepsilon_c)$ instead of the first-order approximation, and check whether every optimized point satisfies the Chernoff-bound condition $x \ge 0.5$; a visible gap or a violated condition would mean the reported RSMA gains do not correspond to the true maximum of the successful computation probability.

Watch

Extended reading notes

Core claim

The central claim is that in an uplink RSMA-aided MEC system with finite-blocklength constraints, splitting one user's message into two streams and decoding them in the order $s_{1,1} \rightarrow s_2 \rightarrow s_{1,2}$ yields a higher Successful Computation Probability than conventional NOMA with decoding order $s_1 \rightarrow s_2$, and achieves the same SCP at a shorter blocklength. The paper derives the SCP as the product of the successful offloading probability and the successful execution probability, approximates the offloading error probability using the finite-blocklength normal approximation with channel dispersion, and then maximizes a Chernoff-bound proxy of the SCP by alternately optimizing the offloading factor, the RSMA power allocation, and the task-splitting factor. Numerical results over Rayleigh fading show RSMA outperforming NOMA across blocklengths, task sizes, and signal-to-noise ratios, with the gain saturating at large blocklengths.

Load-bearing premise

The whole argument rests on the assumption that the per-stream decoding error probabilities $\varepsilon_a,\varepsilon_b,\varepsilon_c$ stay small enough that products such as $\varepsilon_a\varepsilon_b$ can be dropped, and that the Chernoff bound's validity condition $x \ge 0.5$ holds at every optimized point; if those fail in the low-success regimes the paper plots, the optimized decisions may not actually maximize the successful computation probability.

Editorial extensions

If this is right

  • In a two-user MEC offloading scenario with finite-blocklength links, RSMA with the $s_{1,1} \rightarrow s_2 \rightarrow s_{1,2}$ decoding order attains a higher SCP than NOMA at the same blocklength and signal-to-noise ratio.
  • RSMA reaches the same SCP as NOMA at a shorter blocklength, meaning lower offloading latency for the same reliability in short-packet edge computing.
  • The RSMA advantage grows with task size and saturates as blocklength becomes large, so the benefit is concentrated in the short-packet regime where MEC latency targets are tight.
  • The closed-form offloading factor $\lambda_k^* = \max\{0,\, 1 - (T - N T_s) f_{\text{user}}/(M_k C_{\text{cpu}})\}$ determines how much each user must offload given the blocklength and time budget.
  • The proposed alternating optimization with successive convex approximation converges with complexity $O\big((K{+}1)^{3.5}\log(1/\tau)\big)$ and can be extended in principle to $K$ users and MIMO with revised SCP expressions.

Reading between the lines

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

  • Because NOMA is a special case of RSMA in this setup (setting $P_{1,1}$ to $P_1$ or $0$), the reported gain can be attributed to the extra degree of freedom from splitting one user's stream; a natural extension is to quantify the gain from splitting both users or choosing which user to split adaptively per channel realization.
  • The Chernoff-bound proxy suggests the optimization may be conservative or mis-tuned in low-SCP regimes; optimizing the exact product form instead could change the shape of the RSMA-versus-NOMA gap in the very short-blocklength region.
  • The latency benefit implies a direct reliability-delay trade-off; a testable extension is to formulate a joint SCP-latency Pareto problem rather than fixing the time budget and minimizing blocklength.
  • The framework assumes perfect channel state information and no queueing at the edge server; extending to imperfect channel state information would likely shrink but not eliminate the RSMA advantage, since the SINR-balancing mechanism is the core of the gain.
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

4 major / 5 minor

Summary. The paper studies a two-user uplink rate-splitting multiple access (RSMA) system in which users offload computation tasks to a mobile edge computing (MEC) server under finite-blocklength (FBL) constraints. It derives an expression for the successful computation probability (SCP), which is the probability that all offloaded and local tasks are completed within a delay budget, and then formulates a joint optimization of the offloading factor, transmit power allocation, and task-splitting factor. The problem is solved by alternating optimization (AO) with successive convex approximation (SCA) for the power and splitting subproblems, and a closed-form offloading factor is proposed. Numerical results compare RSMA with non-orthogonal multiple access (NOMA) and report that RSMA achieves a higher SCP and can reach the same SCP with shorter blocklengths, i.e., lower latency.

Significance. If the numerical claims are validated, the paper provides a useful extension of RSMA to short-packet mobile edge computing, combining a finite-blocklength reliability analysis with an iterative resource-allocation algorithm. A strength of the paper is that the system model and approximations are explicit, and the algorithm is described in sufficient detail to be reproduced. However, because NOMA is a special case of RSMA (Remark 1), the qualitative ranking RSMA-vs-NOMA is structurally expected; the contribution rests on the quantitative SCP analysis and on the reliability of the optimization. The main risks are the unvalidated surrogate objective used in the optimization and the direction of the convergence proof, both of which affect the reported numerical gains.

major comments (4)
  1. [Section III-A, Eqs. (14)-(18), and Section V] The optimization does not directly maximize the true SCP in Eq. (15) but a linearized surrogate 1-(2εa+2εb+εc) in Eq. (18), after dropping all product terms such as εaεb and ε1ε2. The paper does not verify that these terms are negligible in the operating region of Figs. 2-4. In the highlighted point of Fig. 2(b) (SCP≈0.6-0.8), per-stream error probabilities are of order 0.1-0.3, so the dropped product terms contribute roughly 0.01-0.1 to Ps, which is comparable to the reported RSMA-NOMA gap. Please quantify the approximation error, state explicitly whether the plotted SCP curves are computed from the true expression (15)-(16) using the optimized variables or from the surrogate (18)/(21), and if the latter, recompute the curves with the true objective.
  2. [Section IV, Eqs. (19)-(21)] The transformation from Problem (19) to Problem (21) uses the bound exp(-x^2/2) ≥ Q(x) for x≥0.5, but the algorithm never enforces f(γi,Mi) ≥ 0.5 (or even f≥0). When f(γi,Mi)<0, the inequality reverses, and exp(-f^2/2) can be much smaller than Q(f); minimizing the exponential then rewards large |f| regardless of sign, potentially favoring negative f with large true error. The paper reports operating points with SCP=0, e.g., Fig. 2(a), which can correspond to f<0 for some streams. In those regimes Problem (21) is not an upper bound on Problem (19), so the optimized powers and splitting factors are not guaranteed to minimize the true error. Please enforce the condition f_i≥0.5 as a constraint, or verify after optimization that it holds at the returned solutions and restrict the operating regime accordingly.
  3. [Section IV-D, Eqs. (43)-(45)] The convergence proof has the wrong monotonicity direction. Problem (21) and its subproblems are minimizations, so solving the nth subproblem should produce an objective value no larger than the value at the previous iterate; Eq. (43) asserts the opposite inequality. Consequently Eq. (45) establishes that the sequence of the quantity being minimized is non-decreasing, which is directionally inconsistent with the goal of minimizing the error and does not prove convergence to a local minimum. Please correct the inequality chain (it should be non-increasing for a minimization algorithm) and state the resulting stationarity claim for Algorithm 1.
  4. [Section IV-A, Lemma 1 and Appendix] The closed-form offloading factor in Eq. (25) relies on the claim that the objective is monotonically increasing in λ. The proof in the Appendix, around Eq. (47), assumes log(1+γi) - λiMi/N ≥ 0, i.e., that the FBL rate is not above the corresponding capacity. This is the same 'good' regime required by the Chernoff step; in the low-SCP regime this condition can fail and the derivative can change sign, so the closed-form λ* may not minimize the objective. In addition, Lemma 1 enforces only the local computation constraints (23a), not the MEC computation constraint (23b), so the returned λ can violate feasibility if the server computation time is the bottleneck. Please state the regime required for Lemma 1 and check constraint (23b) explicitly, or justify that the lower-bound λ always satisfies it in the simulated parameters.
minor comments (5)
  1. [Section IV-C, Eq. (38) vs Eq. (42)] Problem (38) is written as a maximization of the objective 2exp(-ta)+2exp(-tb)+exp(-tc), whereas the equivalent transformed problem (42) is written as a minimization of the same objective. Please make the sign consistent.
  2. [Section IV-B, Eq. (34a)] In Eq. (34a), the first term on the left-hand side appears as 't1/N'; it should be t_i/N to match the derivation in Eqs. (28)-(30).
  3. [Figure 2 caption] The last panel is labeled '(c) N=1000' but should be labeled '(d) N=1000'; panel (c) appears twice.
  4. [Section IV-D, complexity discussion] The complexity claim uses X=K+1 as the number of variables, but Problem (37) includes the slack variables ρ, t, and t1 in addition to the powers m; please clarify how these auxiliary variables are counted in the SOCP complexity.
  5. [Section V] The simulations are averaged over only 100 random channel realizations. Since the quoted differences in SCP between RSMA and NOMA are of order 0.1, please provide confidence intervals or error bars, or increase the number of realizations.

Circularity Check

1 steps flagged · score 6.0 of 10

RSMA-vs-NOMA gain is partially built into the model: NOMA is defined as a special case of RSMA, so the reported SCP/latency advantage is a by-construction consequence; the SCP derivation itself is not circular.

  1. self definitional [Section II-A2 (Remark 1); Section VI (Conclusion)]
    "If P1,1 in RSMA scheme is set to P1 or 0, RSMA boils down to NOMA with decoding order of s1 → s2 or s2 → s1, respectively. Since 0≤ P1,1≤ P1, we can regard NOMA as a subset of RSMA. In conclusion, RSMA continuously produces a higher SCP than NOMA in FBL regime, allowing RSMA to provide a more dependable performance."

    In Problems (17)/(21), the RSMA feasible set contains the NOMA feasible set via P1,1=P1, β=1 (or P1,1=0, β=0), with the same SCP objective in (15)/(18). Thus the optimal SCP over RSMA dominates the optimal SCP over NOMA by construction, and the 'improve the SCP' and 'lower latency' conclusions follow directly from this inclusion rather than from an independent empirical test. The simulations quantify the gap, but the direction of the advantage is built into the model definition; at best the magnitude is a numerical output, while the claimed superiority is not a discovered prediction.

full rationale

The one structural circularity is the RSMA-vs-NOMA comparison. Remark 1 makes NOMA a special case of RSMA, so the RSMA feasible set contains the NOMA feasible set; with the same SCP objective, the model-level RSMA optimum dominates NOMA's. The abstract and conclusion present this dominance as the main finding, but its direction is by construction. That said, the SCP derivation (Eqs. (11)-(15)) and the AO/SCA optimization are self-contained: no parameter is fitted to a subset of data and then used to predict a closely related quantity, so there is no fitted-input-as-prediction circularity. The self-citation [26] is used to select the decoding order, but it is a published, parameter-free prior analysis rather than an unverified uniqueness claim, so it is not separately counted as load-bearing circularity. Separate correctness risks, which are not circularity, include: Eq. (18) drops product terms such as εaεb that may be non-negligible in the plotted SCP regime; Eq. (20)'s Chernoff bound requires x≥0.5, a condition not enforced in Problem (21); and the convergence proof in Eq. (45) establishes a non-decreasing objective sequence for a minimization problem, which is the wrong direction. These issues affect the reliability of the numerical magnitudes and the convergence claim, but they do not change the circularity assessment of the core comparison.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper introduces no fitted constants or new physical entities. The central claim rests on standard FBL approximations, several domain simplifications, and an upper-bound transformation of the objective. The most influential modeling choice is the SCP approximation and the Chernoff bound, which are used both in the optimization and in the simulations, making the numerical results conditional on their validity.

assumptions (6)
  • domain assumption Finite-blocklength rate approximation: r ≈ C(γ) - sqrt(V/N) Q^{-1}(ε) holds for the considered blocklengths and error probabilities.
    Invoked in Eq. (11) from [33] and used throughout Section III-A to express per-stream error probabilities. The approximation is standard but only accurate for sufficiently large N.
  • ad hoc to paper The joint success probability can be approximated by dropping higher-order error terms: (1-ε1)(1-ε2) ≈ 1 - 2εa - 2εb - εc.
    Eq. (14) and (18) in Section III-A. Valid only when per-stream error probabilities are small, which is not guaranteed in the simulated low-SCP regimes.
  • ad hoc to paper Chernoff bound exp(-x^2/2) ≥ Q(x) is applied for all streams, with the condition x ≥ 0.5 not enforced.
    Eq. (20)-(21) in Section IV. The bound is not valid for negative f(γ,M), which can occur when the coding rate exceeds capacity.
  • domain assumption Perfect CSIT/CSIR and Rayleigh fading with known channel statistics.
    Section II-A1. The SINR expressions and SCP analysis assume the server perfectly knows the channels.
  • domain assumption No queue delay at the MEC server, t1 and t4 are negligible, and tasks are arbitrarily divisible.
    Section II. These simplify the timing constraint t2+t3 ≤ T and the offloading factor λ being continuous in [0,1].
  • domain assumption Decoding order s1,1 → s2 → s1,2 is fixed and is the one that yields the RSMA gain.
    Section II-A1, based on the authors' prior work [26]. The gains reported are conditional on this order and on the absence of SIC error propagation beyond the modeled cascade.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Uplink Rate-Splitting Multiple Access for Mobile Edge Computing with Short-Packet Communications." pith.science (2026). https://pith.science/paper/2NXR6Y3V

@misc{pith2026250204827,
  author       = {Pith},
  title        = {Pith review of: Uplink Rate-Splitting Multiple Access for Mobile Edge Computing with Short-Packet Communications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2NXR6Y3V}},
  note         = {Machine review of arXiv:2502.04827}
}
read the original abstract

In this paper, a Rate-Splitting Multiple Access (RSMA) scheme is proposed to assist a Mobile Edge Computing (MEC) system where local computation tasks from two users are offloaded to the MEC server, facilitated by uplink RSMA for processing. The efficiency of the MEC service is hence primarily influenced by the RSMA-aided task offloading phase and the subsequent task computation phase, where reliable and low-latency communication is required. For this practical consideration, short-packet communication in the Finite Blocklength (FBL) regime is introduced. In this context, we propose a novel uplink RSMA-aided MEC framework and derive the overall Successful Computation Probability (SCP) with FBL consideration. To maximize the SCP of our proposed RSMA-aided MEC, we strategically optimize: (1) the task offloading factor which determines the number of tasks to be offloaded and processed by the MEC server; (2) the transmit power allocation between different RSMA streams; and (3) the task-splitting factor which decides how many tasks are allocated to splitting streams, while adhering to FBL constraints. To address the strong coupling between these variables in the SCP expression, we apply the Alternative Optimization method, which formulates tractable subproblems to optimize each variable iteratively. The resultant non-convex subproblems are then tackled by Successive Convex Approximation. Numerical results demonstrate that applying uplink RSMA in the MEC system with FBL constraints can not only improve the SCP performance but also provide lower latency in comparison to conventional transmission scheme such as Non-orthogonal Multiple Access (NOMA).

Figures

Figures reproduced from arXiv: 2502.04827 by the authors.

Figure 1
Figure 1. The RSMA-aided MEC system with two users. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The successful computation probability performanc [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. The successful computation probability performanc [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The successful computation probability performanc [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    Network 2030 A Blueprint of Technology, Applicat ions and Mar- ket Drivers Towards the Y ear 2030 and Beyond,

    R. Li, “Network 2030 A Blueprint of Technology, Applicat ions and Mar- ket Drivers Towards the Y ear 2030 and Beyond,” tech. rep., In ternational Telecommunication Union (ITU), 2019

    work page 2019

  2. [2]

    6G Wireless Networks: Vision, Requirements, Ar chitecture, and Key Technologies,

    Z. Zhang, Y . Xiao, Z. Ma, M. Xiao, Z. Ding, X. Lei, G. K. Kara giannidis, and P . Fan, “6G Wireless Networks: Vision, Requirements, Ar chitecture, and Key Technologies,” IEEE V ehicular Technology Magazine, vol. 14, no. 3, pp. 28–41, 2019

    work page 2019

  3. [3]

    A Sur vey on Mobile Edge Computing: The Communication Perspective,

    Y . Mao, C. Y ou, J. Zhang, K. Huang, and K. B. Letaief, “A Sur vey on Mobile Edge Computing: The Communication Perspective,” IEEE Communications Surveys & Tutorials , vol. 19, no. 4, pp. 2322–2358, 2017

    work page 2017

  4. [4]

    Mobile Edge Computing: A Survey on Archi- tecture and Computation Offloading,

    P . Mach and Z. Becvar, “Mobile Edge Computing: A Survey on Archi- tecture and Computation Offloading,” IEEE Communications Surveys & Tutorials, vol. 19, no. 3, pp. 1628–1656, 2017. 11

    work page 2017

  5. [5]

    Mobile E dge Computing: A Survey,

    N. Abbas, Y . Zhang, A. Taherkordi, and T. Skeie, “Mobile E dge Computing: A Survey,” IEEE Internet of Things Journal , vol. 5, no. 1, pp. 450–465, 2017

    work page 2017

  6. [6]

    Mobile-Edge Computing – Introductory Technical White Paper

    ETSI, “Mobile-Edge Computing – Introductory Technical White Paper.” https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge Computing - Introductory Technical White Paper V1%2018-09-14.pdf , Sep. 2014. Accessed at 15/12/2024

    work page 2018

  7. [7]

    Energy-Efficien t Resource Allocation for Mobile-Edge Computation Offloading,

    C. Y ou, K. Huang, H. Chae, and B.-H. Kim, “Energy-Efficien t Resource Allocation for Mobile-Edge Computation Offloading,” IEEE Transac- tions on Wireless Communications , vol. 16, no. 3, pp. 1397–1411, 2016

    work page 2016

  8. [8]

    Energy - Efficient Resource Allocation for Mobile Edge Computing wit h Multiple Relays,

    X. Li, R. Fan, H. Hu, N. Zhang, X. Chen, and A. Meng, “Energy - Efficient Resource Allocation for Mobile Edge Computing wit h Multiple Relays,” IEEE Internet of Things Journal , vol. 9, no. 13, pp. 10732– 10750, 2021

    work page 2021

Show all 46 references
  1. [9]

    Joint Offloading and Computa tion Energy Efficiency Maximization in A Mobile Edge Computing Sy stem,

    H. Sun, F. Zhou, and R. Q. Hu, “Joint Offloading and Computa tion Energy Efficiency Maximization in A Mobile Edge Computing Sy stem,” IEEE Transactions on V ehicular Technology , vol. 68, no. 3, pp. 3052– 3056, 2019

    2019

  2. [10]

    Non-Orthogonal Multiple Access (NOMA) for Cel lular Future Radio Access,

    Y . Saito, Y . Kishiyama, A. Benjebbour, T. Nakamura, A. L i, and K. Higuchi, “Non-Orthogonal Multiple Access (NOMA) for Cel lular Future Radio Access,” in 2013 IEEE 77th vehicular technology confer- ence (VTC Spring) , pp. 1–5, IEEE, 2013

    2013

  3. [11]

    Powe r-Domain Non-Orthogonal Multiple Access (NOMA) in 5G Systems: Poten tials and Challenges,

    S. R. Islam, N. Avazov, O. A. Dobre, and K.-S. Kwak, “Powe r-Domain Non-Orthogonal Multiple Access (NOMA) in 5G Systems: Poten tials and Challenges,” IEEE Communications Surveys & Tutorials , vol. 19, no. 2, pp. 721–742, 2016

    2016

  4. [12]

    Joint Spectral and Energy Efficiency Optimization f or Downlink NOMA Networks,

    W. U. Khan, F. Jameel, T. Ristaniemi, S. Khan, G. A. S. Sid hu, and J. Liu, “Joint Spectral and Energy Efficiency Optimization f or Downlink NOMA Networks,” IEEE Transactions on Cognitive Communications and Networking , vol. 6, no. 2, pp. 645–656, 2019

    2019

  5. [13]

    Spectral Efficiency Optimization for Next Generation NOMA -Enabled IoT Networks,

    W. U. Khan, J. Liu, F. Jameel, V . Sharma, R. J¨ antti, and Z . Han, “Spectral Efficiency Optimization for Next Generation NOMA -Enabled IoT Networks,” IEEE Transactions on V ehicular Technology , vol. 69, no. 12, pp. 15284–15297, 2020

    2020

  6. [14]

    Multi-Antenna NOMA for Comp uta- tion Offloading in Multiuser Mobile Edge Computing Systems,

    F. Wang, J. Xu, and Z. Ding, “Multi-Antenna NOMA for Comp uta- tion Offloading in Multiuser Mobile Edge Computing Systems, ” IEEE Transactions on Communications , vol. 67, no. 3, pp. 2450–2463, 2019

    2019

  7. [15]

    Joint Power an d Time Allocation for NOMA–MEC Offloading,

    Z. Ding, J. Xu, O. A. Dobre, and H. V . Poor, “Joint Power an d Time Allocation for NOMA–MEC Offloading,” IEEE Transactions on V ehicular Technology, vol. 68, no. 6, pp. 6207–6211, 2019

    2019

  8. [16]

    Resource Allocation for Hybrid NOMA MEC Offloading,

    J. Zhu, J. Wang, Y . Huang, F. Fang, K. Navaie, and Z. Ding, “Resource Allocation for Hybrid NOMA MEC Offloading,” IEEE Transactions on Wireless Communications, vol. 19, no. 7, pp. 4964–4977, 2020

    2020

  9. [17]

    Optimal Resource Allocation for Delay Minimization in NOM A-MEC Networks,

    F. Fang, Y . Xu, Z. Ding, C. Shen, M. Peng, and G. K. Karagia nnidis, “Optimal Resource Allocation for Delay Minimization in NOM A-MEC Networks,” IEEE Transactions on Communications , vol. 68, no. 12, pp. 7867–7881, 2020

    2020

  10. [18]

    Delay Min imization for NOMA-MEC Offloading,

    Z. Ding, D. W. K. Ng, R. Schober, and H. V . Poor, “Delay Min imization for NOMA-MEC Offloading,” IEEE Signal Processing Letters , vol. 25, no. 12, pp. 1875–1879, 2018

    2018

  11. [19]

    Enhance Latency-Const rained Com- putation in MEC Networks Using Uplink NOMA,

    Y . Y e, R. Q. Hu, G. Lu, and L. Shi, “Enhance Latency-Const rained Com- putation in MEC Networks Using Uplink NOMA,” IEEE Transactions on Communications , vol. 68, no. 4, pp. 2409–2425, 2020

    2020

  12. [20]

    Reliabili ty-Oriented Design Framework in NOMA-Assisted Mobile Edge Computing,

    Z. Liu, Y . Zhu, Y . Hu, P . Sun, and A. Schmeink, “Reliabili ty-Oriented Design Framework in NOMA-Assisted Mobile Edge Computing,” IEEE Access, vol. 10, pp. 103598–103609, 2022

    2022

  13. [21]

    Rate-Splitting Multiple Access: Fundamentals, Survey, a nd Future Research Trends,

    Y . Mao, O. Dizdar, B. Clerckx, R. Schober, P . Popovski, a nd H. V . Poor, “Rate-Splitting Multiple Access: Fundamentals, Survey, a nd Future Research Trends,” IEEE Communications Surveys & Tutorials , vol. 24, no. 4, pp. 2073–2126, 2022

    2022

  14. [22]

    Rate-Splitting Multip le Access for Downlink Communication Systems: Bridging, Generalizing, and Out- performing SDMA and NOMA,

    Y . Mao, B. Clerckx, and V . O. Li, “Rate-Splitting Multip le Access for Downlink Communication Systems: Bridging, Generalizing, and Out- performing SDMA and NOMA,” Journal on Wireless Communications and Networking , no. 133 (2018), 2018

    2018

  15. [23]

    Max-Min Fairness and PHY-Layer De sign of Uplink MIMO Rate-Splitting Multiple Access with Finite Blo cklength,

    J. Xu and B. Clerckx, “Max-Min Fairness and PHY-Layer De sign of Uplink MIMO Rate-Splitting Multiple Access with Finite Blo cklength,” IEEE Transactions on Communications , pp. 1–1, 2024

    2024

  16. [24]

    Rate-Splittin g Multiple Access With Finite Blocklength for Short-Packet and Low-La tency Downlink Communications,

    Y . Xu, Y . Mao, O. Dizdar, and B. Clerckx, “Rate-Splittin g Multiple Access With Finite Blocklength for Short-Packet and Low-La tency Downlink Communications,” IEEE Transactions on V ehicular Technol- ogy, vol. 71, no. 11, pp. 12333–12337, 2022

    2022

  17. [25]

    A Primer on Rate-Splitting Multiple Access: Tutorial, Myths, and Frequently Asked Questions,

    B. Clerckx, Y . Mao, E. A. Jorswieck, J. Y uan, D. J. Love, E . Erkip, and D. Niyato, “A Primer on Rate-Splitting Multiple Access: Tutorial, Myths, and Frequently Asked Questions,” IEEE Journal on Selected Areas in Communications , vol. 41, no. 5, pp. 1265–1308, 2023

    2023

  18. [26]

    Rate-Splitting Multi ple Access for Short-Packet Uplink Communications: A Finite Blocklength Analysis,

    J. Xu, O. Dizdar, and B. Clerckx, “Rate-Splitting Multi ple Access for Short-Packet Uplink Communications: A Finite Blocklength Analysis,” IEEE Communications Letters , vol. 27, no. 2, pp. 517–521, 2023

    2023

  19. [27]

    Max-Min Fairne ss of Rate- Splitting Multiple Access With Finite Blocklength Communi cations,

    Y . Xu, Y . Mao, O. Dizdar, and B. Clerckx, “Max-Min Fairne ss of Rate- Splitting Multiple Access With Finite Blocklength Communi cations,” IEEE Transactions on V ehicular Technology , vol. 72, no. 5, pp. 6816– 6821, 2023

    2023

  20. [28]

    Network Slicing fo r eMBB, URLLC, and mMTC: An Uplink Rate-Splitting Multiple Access A p- proach,

    Y . Liu, B. Clerckx, and P . Popovski, “Network Slicing fo r eMBB, URLLC, and mMTC: An Uplink Rate-Splitting Multiple Access A p- proach,” IEEE Transactions on Wireless Communications , 2023

    2023

  21. [29]

    Delay Minimization for Rate-Splitting Multiple Access-Based Mu lti-Server MEC Offloading,

    M. Diamanti, C. Pelekis, E. E. Tsiropoulou, and S. Papav assiliou, “Delay Minimization for Rate-Splitting Multiple Access-Based Mu lti-Server MEC Offloading,” IEEE/ACM Transactions on Networking , vol. 32, no. 2, pp. 1035–1047, 2024

    2024

  22. [30]

    Delay Minimiz ation Using Hybrid RSMA-TDMA for Mobile Edge Computing,

    F. Xiao, P . Chen, H. Wu, Y . Mao, and H. Liu, “Delay Minimiz ation Using Hybrid RSMA-TDMA for Mobile Edge Computing,” Electronics, vol. 12, no. 11, p. 2550, 2023

    2023

  23. [31]

    Rate-Splitting Multiple Access Aided Mobile Edge Computi ng With Randomly Deployed Users,

    P . Chen, H. Liu, Y . Y e, L. Y ang, K. J. Kim, and T. A. Tsiftsi s, “Rate-Splitting Multiple Access Aided Mobile Edge Computi ng With Randomly Deployed Users,” IEEE Journal on Selected Areas in Com- munications, vol. 41, no. 5, pp. 1549–1565, 2023

    2023

  24. [32]

    Finite Blockl ength Per- formance of Cooperative Multi-Terminal Wireless Industri al Networks,

    Y . Hu, M. Serror, K. Wehrle, and J. Gross, “Finite Blockl ength Per- formance of Cooperative Multi-Terminal Wireless Industri al Networks,” IEEE Transactions on V ehicular Technology , vol. 67, no. 7, pp. 5778– 5792, 2018

    2018

  25. [33]

    Channel Coding Rate in the Finite Blocklength Regime,

    Y . Polyanskiy, H. V . Poor, and S. V erdu, “Channel Coding Rate in the Finite Blocklength Regime,” IEEE Transactions on Information Theory , vol. 56, no. 5, pp. 2307–2359, 2010

    2010

  26. [34]

    Optimu m Power Con- trol at Finite Blocklength,

    W. Y ang, G. Caire, G. Durisi, and Y . Polyanskiy, “Optimu m Power Con- trol at Finite Blocklength,” IEEE Transactions on Information Theory , vol. 61, pp. 4598–4615, Sep. 2015

    2015

  27. [35]

    Throughput of Cognitive Radi o Systems with Finite Blocklength Codes,

    G. Ozcan and M. C. Gursoy, “Throughput of Cognitive Radi o Systems with Finite Blocklength Codes,” IEEE Journal on Selected Areas in Communications, vol. 31, no. 11, pp. 2541–2554, 2013

    2013

  28. [36]

    Energ y-Efficient Packet Scheduling with Finite Blocklength Codes: Convexit y Analysis and Efficient Algorithms,

    S. Xu, T.-H. Chang, S.-C. Lin, C. Shen, and G. Zhu, “Energ y-Efficient Packet Scheduling with Finite Blocklength Codes: Convexit y Analysis and Efficient Algorithms,” IEEE Transactions on Wireless Communica- tions, vol. 15, no. 8, pp. 5527–5540, 2016

    2016

  29. [37]

    Blocklength-Limited Performance of Relaying Under Quasi-Static Rayleigh Channels,

    Y . Hu, A. Schmeink, and J. Gross, “Blocklength-Limited Performance of Relaying Under Quasi-Static Rayleigh Channels,” IEEE Transactions on Wireless Communications , vol. 15, no. 7, pp. 4548–4558, 2016

    2016

  30. [38]

    Quasi-S tatic Multiple- Antenna Fading Channels at Finite Blocklength,

    W. Y ang, G. Durisi, T. Koch, and Y . Polyanskiy, “Quasi-S tatic Multiple- Antenna Fading Channels at Finite Blocklength,” IEEE Transactions on Information Theory , vol. 60, no. 7, pp. 4232–4265, 2014

    2014

  31. [39]

    The Dispersion of Nearest- Neighbor Decoding for Additive Non-Gaussian Channels,

    J. Scarlett, V . Y . Tan, and G. Durisi, “The Dispersion of Nearest- Neighbor Decoding for Additive Non-Gaussian Channels,” IEEE Trans- actions on Information Theory , vol. 63, no. 1, pp. 81–92, 2016

    2016

  32. [40]

    Delay Per- formance of Wireless Communications with Imperfect CSI and Finite- Length Coding,

    S. Schiessl, H. Al-Zubaidy, M. Skoglund, and J. Gross, “ Delay Per- formance of Wireless Communications with Imperfect CSI and Finite- Length Coding,” IEEE Transactions on Communications, vol. 66, no. 12, pp. 6527–6541, 2018. 12

    2018

  33. [41]

    Optimal P ower Allocation for QoS-Constrained Downlink Multi-User Netwo rks in The Finite Blocklength Regime,

    Y . Hu, M. Ozmen, M. C. Gursoy, and A. Schmeink, “Optimal P ower Allocation for QoS-Constrained Downlink Multi-User Netwo rks in The Finite Blocklength Regime,” IEEE Transactions on Wireless Communi- cations, vol. 17, no. 9, pp. 5827–5840, 2018

    2018

  34. [42]

    Energy Minimiz ation of Mobile Edge Computing Networks with HARQ in The Finite Blocklength Regime,

    Y . Zhu, Y . Hu, A. Schmeink, and J. Gross, “Energy Minimiz ation of Mobile Edge Computing Networks with HARQ in The Finite Blocklength Regime,” IEEE Transactions on Wireless Communications , vol. 21, no. 9, pp. 7105–7120, 2022

    2022

  35. [43]

    Short- Packet Edge Computing Networks With Execution Uncertainty,

    X. Lai, T. Wu, C. Pan, L. Mai, and A. Nallanathan, “Short- Packet Edge Computing Networks With Execution Uncertainty,” IEEE Transactions on Green Communications and Networking , 2024

    2024

  36. [44]

    Offloading Schemes in Mobile Edge Co mputing for Ultra-Reliable Low Latency Communications,

    J. Liu and Q. Zhang, “Offloading Schemes in Mobile Edge Co mputing for Ultra-Reliable Low Latency Communications,” Ieee Access , vol. 6, pp. 12825–12837, 2018

    2018

  37. [45]

    Joint Offloading and C omputing Optimization in Wireless Powered Mobile-Edge Computing Sy stems,

    F. Wang, J. Xu, X. Wang, and S. Cui, “Joint Offloading and C omputing Optimization in Wireless Powered Mobile-Edge Computing Sy stems,” IEEE transactions on wireless communications , vol. 17, no. 3, pp. 1784– 1797, 2017

    2017

  38. [46]

    CVX: Matlab software for di sciplined convex programming,

    M. Grant, S. Boyd, and Y . Y e, “CVX: Matlab software for di sciplined convex programming,” 2009. 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 Offloading Factor 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 SCP RSMA NOMA

    2009

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.