pith. sign in

arxiv: 1907.11580 · v1 · pith:LREE46TZnew · submitted 2019-07-26 · 💻 cs.DC

Edge User Allocation with Dynamic Quality of Service

Phu Lai , Qiang He , Guangming Cui , Xiaoyu Xia , Mohamed Abdelrazek , Feifei Chen , John Hosking , John Grundy
show 1 more author
Yun Yang
This is my paper

Pith reviewed 2026-05-24 15:22 UTC · model grok-4.3

classification 💻 cs.DC
keywords edge computinguser allocationquality of servicequality of experienceoptimizationheuristic
0
0 comments X

The pith

Allowing dynamic QoS levels generalizes edge user allocation to maximize aggregate user QoE.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper extends the standard edge user allocation problem, where fixed QoS must be met to count a user as allocated, by instead letting each user receive one of several possible QoS levels. The new objective becomes maximizing the sum of individual quality-of-experience values rather than simply the count of served users. An exact optimization method is given for the resulting dynamic-QoS problem together with a fast heuristic for instances too large for the exact solver. Real-world traces are used to show that both methods improve total QoE over fixed-QoS baselines. Service providers therefore gain a concrete way to trade off resource use against user satisfaction on the same edge infrastructure.

Core claim

The dynamic QoS EUA problem is defined by allowing each app user to be assigned any of several discrete QoS levels; an optimal method solves for the assignment that maximizes the sum of resulting QoE values while respecting edge-server capacities, and a heuristic produces good solutions quickly for large instances, both shown effective on a real dataset.

What carries the argument

The dynamic QoS edge user allocation model that treats QoS level choice as a decision variable per user inside a resource-constrained maximization of aggregate QoE.

If this is right

  • Service providers obtain higher aggregate user satisfaction by varying QoS per user rather than enforcing a single fixed level.
  • The exact method certifies the maximum attainable QoE for any given set of users and servers.
  • The heuristic scales the approach to realistic city-scale deployments where the exact solver becomes impractical.
  • Both methods outperform prior fixed-QoS and state-of-the-art baselines on measured edge traces.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Allocations could be recomputed periodically as users move, turning the static model into a dynamic online scheduler.
  • Different QoS tiers could be priced separately, linking the optimization directly to revenue.
  • The same formulation might extend to joint optimization of user allocation and edge-server placement.

Load-bearing premise

QoS levels for different users can be chosen independently without additional unstated limits on total capacity, latency, or fairness that would invalidate the maximization.

What would settle it

On the same dataset and server capacities, the optimal dynamic-QoS solution yields no higher total QoE than the best fixed-QoS allocation.

Figures

Figures reproduced from arXiv: 1907.11580 by Feifei Chen, Guangming Cui, John Grundy, John Hosking, Mohamed Abdelrazek, Phu Lai, Qiang He, Xiaoyu Xia, Yun Yang.

Figure 1
Figure 1. Figure 1: Quality of Experience - Quality of Service correlation [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Dynamic QoS EUA example scenario h5, 7, 6, 6i units of hCP U, RAM, storage, bandwidthi, respectively. Players’ cor￾responding QoE, measured based on Eq. 3, are 1.6, 4.09, and 4.99, respectively. If the server’s available resources are not limited then all players will be able to enjoy the highest QoS level. However, a typical edge server has relatively limited resources so not everyone will be assigned W3.… view at source ↗
Figure 3
Figure 3. Figure 3: Experiment set #1 results (a) Total QoE (b) Elapsed CPU time [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Experiment set #2 results number of users continues to increase while the amount of available resources is fixed, the computing resource for each user becomes more scarce, making Heuristic no longer suitable in these situations. In fact, from 700 users onwards, Heuristic starts being outperformed by Random and VSVBP. Due to being a greedy heuristic, Heuristic always tries to exhaust the edge servers’ resou… view at source ↗
Figure 5
Figure 5. Figure 5: Experiment set #3 results in experiment set 1 ( [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
read the original abstract

In edge computing, edge servers are placed in close proximity to end-users. App vendors can deploy their services on edge servers to reduce network latency experienced by their app users. The edge user allocation (EUA) problem challenges service providers with the objective to maximize the number of allocated app users with hired computing resources on edge servers while ensuring their fixed quality of service (QoS), e.g., the amount of computing resources allocated to an app user. In this paper, we take a step forward to consider dynamic QoS levels for app users, which generalizes but further complicates the EUA problem, turning it into a dynamic QoS EUA problem. This enables flexible levels of quality of experience (QoE) for app users. We propose an optimal approach for finding a solution that maximizes app users' overall QoE. We also propose a heuristic approach for quickly finding sub-optimal solutions to large-scale instances of the dynamic QoS EUA problem. Experiments are conducted on a real-world dataset to demonstrate the effectiveness and efficiency of our approaches against a baseline approach and the state of the art.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript formulates the dynamic QoS edge user allocation (EUA) problem as a generalization of prior EUA work, allowing per-user QoS levels to be chosen dynamically. It presents an optimal (ILP-based) method claimed to maximize aggregate QoE, a heuristic for scalability, and experimental comparisons on a real-world dataset against a baseline and prior state-of-the-art approaches.

Significance. If the ILP correctly encodes all relevant resource-coupling constraints and the experiments demonstrate reliable gains, the work supplies a practically useful extension of EUA that supports flexible QoE provisioning in edge environments. The real-dataset evaluation is a positive element.

major comments (2)
  1. [§3] §3 (problem formulation and ILP): the model must be shown to include explicit server-capacity, latency, and non-overallocation constraints; without them the per-user QoS independence assumption risks producing infeasible allocations that invalidate the claimed QoE maximization.
  2. [§5] §5 (experiments): no complexity analysis or runtime scaling results are provided for the optimal solver on the largest instances, which is load-bearing for the claim that the optimal method is viable alongside the heuristic.
minor comments (2)
  1. Add error bars or statistical significance tests to all reported QoE and allocation metrics in the experimental tables/figures.
  2. [§2] Clarify the exact mapping from chosen QoS level to QoE value (utility function) and whether it is linear or piecewise.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. We address each major comment below and outline the revisions we will make.

read point-by-point responses

  1. Referee: [§3] §3 (problem formulation and ILP): the model must be shown to include explicit server-capacity, latency, and non-overallocation constraints; without them the per-user QoS independence assumption risks producing infeasible allocations that invalidate the claimed QoE maximization.

    Authors: We agree that explicit presentation of these constraints is necessary for clarity. The ILP in Section 3 already encodes server capacity (constraint 3), per-user latency bounds (constraint 4), and non-overallocation (constraint 5). To address the comment, we will insert a dedicated paragraph immediately after the ILP formulation that lists and justifies each of these constraints, removing any ambiguity about feasibility. revision: yes

  2. Referee: [§5] §5 (experiments): no complexity analysis or runtime scaling results are provided for the optimal solver on the largest instances, which is load-bearing for the claim that the optimal method is viable alongside the heuristic.

    Authors: We accept this point. The current experiments focus on solution quality but omit scaling data for the ILP solver. In the revised manuscript we will add a new subsection (5.4) that reports (i) the theoretical complexity of the ILP, (ii) wall-clock runtimes on instances up to the largest size used in the paper, and (iii) a direct comparison of solver time versus the heuristic. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper formulates the dynamic QoS edge user allocation problem as a standard optimization task (maximizing aggregate QoE under per-user QoS levels) and proposes an exact optimal solver plus a heuristic. No equations, fitted parameters, or load-bearing steps are shown that reduce by construction to self-referential quantities, self-citations, or imported ansatzes. The derivation chain consists of a conventional ILP-style model plus approximation and is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; all modeling assumptions remain implicit.

pith-pipeline@v0.9.0 · 5737 in / 962 out tokens · 25036 ms · 2026-05-24T15:22:35.811192+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages · 1 internal anchor

  1. [1]

    In: 2016 23rd International Conference on Telecommunications (ICT)

    Aazam, M., St-Hilaire, M., Lung, C.H., Lambadaris, I.: Mefore: Qoe based resource estimation at fog to enhance qos in iot. In: 2016 23rd International Conference on Telecommunications (ICT). pp. 1–5. IEEE (2016)

    work page 2016

  2. [2]

    Survey on QoE\QoS Correlation Models For Multimedia Services

    Alreshoodi, M., Woods, J.: Survey on qoe \qos correlation models for multimedia services. arXiv preprint arXiv:1306.0221 (2013)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  3. [3]

    In: Proceedings of the first edition of the MCC workshop on Mobile cloud computing

    Bonomi, F., Milito, R., Zhu, J., Addepalli, S.: Fog computing and its role in the internet of things. In: Proceedings of the first edition of the MCC workshop on Mobile cloud computing. pp. 13–16. ACM (2012)

    work page 2012

  4. [4]

    Ericsson, Stockholm (2018), https://www.ericsson.com/en/mobility-report/reports/november-2018

    Cerwall, P., Lundvall, A., Jonsson, P., Carson, S., M¨ oller, R., Jonsson, P., Car- son, S., Lindberg, P., ¨Ohman, K., Sorlie, I., Queir´ os, R., Muller, F., Englund, L., Arvedson, M., Carlsson, A.: Ericsson mobility report. Ericsson, Stockholm (2018), https://www.ericsson.com/en/mobility-report/reports/november-2018

    work page 2018

  5. [5]

    IEEE Network 29(2), 32–38 (2015)

    Chen, M., Zhang, Y., Li, Y., Mao, S., Leung, V.C.: Emc: Emotion-aware mobile cloud computing in 5g. IEEE Network 29(2), 32–38 (2015)

    work page 2015

  6. [6]

    IEEE Transactions on Parallel and Distributed Systems 26(4), 974–983 (2015)

    Chen, X.: Decentralized computation offloading game for mobile cloud computing. IEEE Transactions on Parallel and Distributed Systems 26(4), 974–983 (2015)

    work page 2015

  7. [7]

    In: 2008 4th International Conference on Wireless Communications, Networking and Mobile Computing

    Ding, B., Chen, L., Chen, D., Yuan, H.: Application of rtls in warehouse manage- ment based on rfid and wi-fi. In: 2008 4th International Conference on Wireless Communications, Networking and Mobile Computing. pp. 1–5. IEEE (2008)

    work page 2008

  8. [8]

    IEEE Network 24(2), 36–41 (2010)

    Fiedler, M., Hossfeld, T., Tran-Gia, P.: A generic quantitative relationship between quality of experience and quality of service. IEEE Network 24(2), 36–41 (2010)

    work page 2010

  9. [9]

    Garey, M.R., Johnson, D.S.: Computers and intractability, vol. 29. wh freeman New York (2002)

    work page 2002

  10. [10]

    IEEE/ACM Transactions on Networking (TON) 15(6), 1240–1253 (2007)

    Hande, P., Zhang, S., Chiang, M.: Distributed rate allocation for inelastic flows. IEEE/ACM Transactions on Networking (TON) 15(6), 1240–1253 (2007)

    work page 2007

  11. [11]

    IEEE Transactions on Circuits and Systems for Video Technology 24(4), 669–680 (2013)

    He, J., Wen, Y., Huang, J., Wu, D.: On the cost–qoe tradeoff for cloud-based video streaming under amazon ec2’s pricing models. IEEE Transactions on Circuits and Systems for Video Technology 24(4), 669–680 (2013)

    work page 2013

  12. [12]

    IEEE MultiMedia 24(4), 80–90 (2017)

    Hemmati, M., McCormick, B., Shirmohammadi, S.: Qoe-aware bandwidth alloca- tion for video traffic using sigmoidal programming. IEEE MultiMedia 24(4), 80–90 (2017)

    work page 2017

  13. [13]

    IEEE Communications Magazine 50(4), 28–36 (2012)

    Hobfeld, T., Schatz, R., Varela, M., Timmerer, C.: Challenges of qoe management for cloud applications. IEEE Communications Magazine 50(4), 28–36 (2012)

    work page 2012

  14. [14]

    In: 2016 13th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON)

    Hong, S.T., Kim, H.: Qoe-aware computation offloading scheduling to capture energy-latency tradeoff in mobile clouds. In: 2016 13th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON). pp. 1–9. IEEE (2016)

    work page 2016

  15. [15]

    In: 2011 IEEE International Symposium on Multimedia

    Hoßfeld, T., Seufert, M., Hirth, M., Zinner, T., Tran-Gia, P., Schatz, R.: Quantifi- cation of youtube qoe via crowdsourcing. In: 2011 IEEE International Symposium on Multimedia. pp. 494–499. IEEE (2011)

    work page 2011

  16. [16]

    ETSI white paper 11(11), 1–16 (2015)

    Hu, Y.C., Patel, M., Sabella, D., Sprecher, N., Young, V.: Mobile edge comput- ing—a key technology towards 5g. ETSI white paper 11(11), 1–16 (2015)

    work page 2015

  17. [17]

    In: Image Quality and System Per- formance XII

    Lachat, A., Gicquel, J.C., Fournier, J.: How perception of ultra-high definition is modified by viewing distance and screen size. In: Image Quality and System Per- formance XII. vol. 9396, p. 93960Y. International Society for Optics and Photonics (2015)

    work page 2015

  18. [18]

    Lai et al

    Lai, P., He, Q., Abdelrazek, M., Chen, F., Hosking, J., Grundy, J., Yang, Y.: Optimal edge user allocation in edge computing with variable sized vector bin 16 P. Lai et al. packing. In: International Conference on Service-Oriented Computing. pp. 230–

    work page

  19. [19]

    Journal of Parallel and Distributed Computing (2018)

    Mahmud, R., Srirama, S.N., Ramamohanarao, K., Buyya, R.: Quality of experience (qoe)-aware placement of applications in fog computing environments. Journal of Parallel and Distributed Computing (2018)

    work page 2018

  20. [20]

    IEEE Journal on selected areas in communications 13(7), 1176–1188 (1995)

    Shenker, S.: Fundamental design issues for the future internet. IEEE Journal on selected areas in communications 13(7), 1176–1188 (1995)

    work page 1995

  21. [21]

    In: Computers and communications (ISCC), 2012 IEEE symposium on

    Soyata, T., Muraleedharan, R., Funai, C., Kwon, M., Heinzelman, W.: Cloud- vision: Real-time face recognition using a mobile-cloudlet-cloud acceleration archi- tecture. In: Computers and communications (ISCC), 2012 IEEE symposium on. pp. 59–66. IEEE (2012)

    work page 2012

  22. [22]

    IEEE Transactions on Multimedia 18(8), 1650– 1660 (2016)

    Su, Z., Xu, Q., Fei, M., Dong, M.: Game theoretic resource allocation in media cloud with mobile social users. IEEE Transactions on Multimedia 18(8), 1650– 1660 (2016)

    work page 2016