pith. sign in

arxiv: 1906.08370 · v1 · pith:Y4WAXWUBnew · submitted 2019-06-19 · 💻 cs.NI

Driving Path Stability in VANETs

Mohammed Laroui , Akrem Sellami , Boubakr Nour , Hassine Moungla , Hossam Afifi , Sofiane B.Hacene This is my paper

Pith reviewed 2026-05-25 19:37 UTC · model grok-4.3

classification 💻 cs.NI
keywords VANETlink predictionSupport Vector Regressionmobility predictionrouting stabilitymachine learning
0
0 comments X

The pith

Support Vector Regression predicts vehicle movements in VANETs to select stable communication links in advance.

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

The paper proposes linear regression and support vector regression to forecast future positions of vehicles so that routing decisions can favor links expected to remain connected. Predictions are generated from mobility traces and then compared against actual recorded movements in multiple scenarios. Effectiveness is measured through error rates, with the SVR approach also benchmarked directly against Lagrange interpolation on the same data.

Core claim

SVR allows predicting the movements of the vehicles in the network which gives us a decision for the link state at a future time. Performance is studied by comparing the generated prediction values against real movement traces of different vehicles in various mobility scenarios, with effectiveness shown by the calculated error rate and by comparison to Lagrange interpolation.

What carries the argument

Support Vector Regression (SVR) models vehicle trajectories from position data to output future link-state predictions for routing.

If this is right

  • Routing protocols can drop candidate links whose predicted future state indicates breakage.
  • Link-state decisions become available before the link actually fails.
  • The same SVR model can be retrained on additional traces from the tested scenarios.
  • Error-rate metrics provide a direct numeric basis for comparing prediction methods in VANET routing.

Where Pith is reading between the lines

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

  • Integration into existing VANET simulators would allow direct measurement of end-to-end packet delivery gains.
  • Retraining frequency would need to match the rate at which traffic patterns shift in real deployments.
  • The approach leaves open whether SVR predictions remain accurate when vehicles follow non-standard routes or sudden maneuvers.

Load-bearing premise

The mobility scenarios and real movement traces used for testing are representative enough that the model will perform similarly on unseen traffic conditions.

What would settle it

A new mobility trace where the SVR error rate exceeds the Lagrange interpolation error rate for the same prediction horizon would contradict the reported superiority.

Figures

Figures reproduced from arXiv: 1906.08370 by Akrem Sellami, Boubakr Nour, Hassine Moungla, Hossam Afifi, Mohammed Laroui, Sofiane B.Hacene.

Figure 1
Figure 1. Figure 1: Simulation Scenarios and that the trajectory can be presented with many stochastic models because we can observe several deviations and sudden events in a trajectory, we propose to use four models SVR, adjusted beforehand with the best combination of parameters (according to the trend of the trajectory and its speed history) to improve the prediction results of the trajectory: -SVR1: Used when vehicle spee… view at source ↗
Figure 2
Figure 2. Figure 2: Predicted traces using Lagrange, SVR, and LR [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: SVR, Lagrange & LR Performance Analysis (MAE), root-mean-square error (RMSE), and mean absolute percentage error (MAPE). These evaluation criteria can be defined as : MSE = 1 n Xn i=1 (Pi − Pˆ i) 2 (13) MAE = 1 n Xn i=1 |(Pi − Pˆ i)| (14) RMSE = ( 1 n Xn i=1 (Pi − Pˆ i) 2 ) 1/2 (15) MAP E = 100 n Xn i=1 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
read the original abstract

Vehicular Ad Hoc Network has attracted both research and industrial community due to its benefits in facilitating human life and enhancing the security and comfort. However, various issues have been faced in such networks such as information security, routing reliability, dynamic high mobility of vehicles, that influence the stability of communication. To overcome this issue, it is necessary to increase the routing protocols performances, by keeping only the stable path during the communication. The effective solutions that have been investigated in the literature are based on the link prediction to avoid broken links. In this paper, we propose a new solution based on machine learning concept for link prediction, using LR and Support Vector Regression (SVR) which is a variant of the Support Vector Machine (SVM) algorithm. SVR allows predicting the movements of the vehicles in the network which gives us a decision for the link state at a future time. We study the performance of SVR by comparing the generated prediction values against real movement traces of different vehicles in various mobility scenarios, and to show the effectiveness of the proposed method, we calculate the error rate. Finally, we compare this new SVR method with Lagrange interpolation solution.

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 / 0 minor

Summary. The paper proposes using Linear Regression and Support Vector Regression (SVR) to predict vehicle movements in VANETs, enabling future link-state decisions that support more stable routing paths. Performance is assessed by comparing SVR-generated predictions to real movement traces across mobility scenarios, reporting error rates, and benchmarking against Lagrange interpolation.

Significance. If the central claim holds, the work could offer a data-driven alternative for link prediction in high-mobility vehicular networks. The use of real traces for validation is a constructive element. However, because the evaluation stops at scalar prediction error and does not demonstrate routing-level gains, the practical significance for path stability remains unestablished.

major comments (2)
  1. [Abstract] Abstract: The stated goal is to improve routing reliability by maintaining stable paths via link prediction, yet the evaluation reports only position/velocity prediction error rates versus real traces and Lagrange interpolation; no routing protocol is instrumented and no link-duration, path-stability, PDR, or route-lifetime metrics are computed, so the error reduction does not establish the claimed routing benefit.
  2. [Abstract] Abstract: No information is supplied on training-set size, feature selection, SVR hyper-parameter tuning, or statistical significance testing of the error-rate comparisons, preventing assessment of whether the reported improvements are robust or generalizable.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the scope of our evaluation and the need for additional methodological details. We respond to each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The stated goal is to improve routing reliability by maintaining stable paths via link prediction, yet the evaluation reports only position/velocity prediction error rates versus real traces and Lagrange interpolation; no routing protocol is instrumented and no link-duration, path-stability, PDR, or route-lifetime metrics are computed, so the error reduction does not establish the claimed routing benefit.

    Authors: We agree that the manuscript evaluates prediction accuracy (position/velocity error versus real traces and Lagrange interpolation) rather than instrumenting a routing protocol or reporting metrics such as link duration, PDR, or route lifetime. The core contribution is showing that SVR yields lower prediction errors than the baseline, which we position as an enabling step for more reliable future link-state decisions. We will revise the abstract and introduction to explicitly limit the claims to the prediction task and its potential implications for path stability, without asserting quantified routing gains. revision: partial

  2. Referee: [Abstract] Abstract: No information is supplied on training-set size, feature selection, SVR hyper-parameter tuning, or statistical significance testing of the error-rate comparisons, preventing assessment of whether the reported improvements are robust or generalizable.

    Authors: We acknowledge the omission of these reproducibility details. The revised version will add: (i) the number of samples and time windows used for training from each mobility trace, (ii) the input features (e.g., coordinates, velocity, time), (iii) the hyper-parameter selection procedure for SVR (kernel, C, epsilon, via cross-validation), and (iv) basic statistical measures (mean error and variance across scenarios) to support the reported comparisons. revision: yes

Circularity Check

0 steps flagged

No circularity; SVR predictions evaluated against external real traces

full rationale

The paper trains SVR (and LR) on vehicle mobility data to predict future positions/velocities, then measures error rates directly against held-out real movement traces from various scenarios and compares the scalar error to Lagrange interpolation. No equation or claim reduces a reported prediction to a fitted parameter by construction, no self-citation is invoked as a uniqueness theorem or load-bearing premise, and the evaluation uses independent external traces rather than re-using the training fit. The derivation chain is therefore self-contained against the stated benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The approach implicitly relies on standard SVR assumptions (kernel choice, regularization) and the representativeness of the mobility traces, none of which are detailed.

pith-pipeline@v0.9.0 · 5743 in / 1105 out tokens · 22695 ms · 2026-05-25T19:37:25.832690+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. [1]

    Wireless network protocol systems and metho ds,

    D. Petite, “Wireless network protocol systems and metho ds,” Google Patents, Jan. 2018, uS Patent 9,860,820

    work page 2018

  2. [2]

    Distributed int erference management in medical wireless sensor networks,

    H. Moungla, K. Haddadi, and S. Boudjit, “Distributed int erference management in medical wireless sensor networks,” in 2016 13th IEEE annual consumer communications & networking conference (C CNC). IEEE, 2016, pp. 151–155

    work page 2016

  3. [3]

    Cost -effective reliability-and energy-based intra-WBAN interference mi tigation,

    H. Moungla, A. Jarray, A. Karmouch, and A. Mehaoua, “Cost -effective reliability-and energy-based intra-WBAN interference mi tigation,” in 2014 IEEE Global Communications Conference. IEEE, 2014, pp. 2399– 2404

    work page 2014

  4. [4]

    G. L. St¨ uber, Principles of mobile communication . Springer, 2017, vol. 3. MSE MAE RMSE MAPE Scenarios LR SVR Lagrange LR SVR Lagrange LR SVR Lagrange LR SVR Lagrange City 9.9514 7.2129 2.8095e-04 3.1028 2.0055 0.0138 3.5410 2.6857 0.0168 0.0921 0.0324 3.7166e-04 Road 4.1247 3.8200 0.0011 2.8751 1.2591 0.0150 2.0158 1.9545 0.0331 0.4216 0.2233 0.0022 In...

    work page 2017

  5. [5]

    J. Loo, J. L. Mauri, and J. H. Ortiz, Mobile ad hoc networks: current status and future trends . CRC Press, 2016

    work page 2016

  6. [6]

    A comprehensive survey on vehicular ad hoc network,

    S. Al-Sultan, M. M. Al-Doori, A. H. Al-Bayatti, and H. Zed an, “A comprehensive survey on vehicular ad hoc network,” Journal of network and computer applications , vol. 37, pp. 380–392, 2014

    work page 2014

  7. [7]

    A reliable, effici ent routing protocol for dynamic topology in Wireless Body Area Network s using min-max multi-commodity flow model,

    H. Moungla, N. Touati, and A. Mehaoua, “A reliable, effici ent routing protocol for dynamic topology in Wireless Body Area Network s using min-max multi-commodity flow model,” in 2012 IEEE 14th Interna- tional Conference on e-Health Networking, Applications an d Services (Healthcom). IEEE, 2012, pp. 470–473

    work page 2012

  8. [8]

    Performance comparison of aodv, ds dv, olsr and dsr routing protocols in mobile ad hoc networks,

    S. Ade and P . Tijare, “Performance comparison of aodv, ds dv, olsr and dsr routing protocols in mobile ad hoc networks,” International journal of information technology and knowledge management , vol. 2, no. 2, pp. 545–548, 2010

    work page 2010

  9. [9]

    A min-ma x multi- commodity flow model for wireless body area networks routing ,

    H. Moungla, N. Touati, O. Salem, and A. Mehaoua, “A min-ma x multi- commodity flow model for wireless body area networks routing ,” IEEE Consumer Communications and Networking Conference (CCNC) , pp. 424–428, 2012

    work page 2012

  10. [10]

    V ehicu lar cloud networks: Challenges, architectures, and future directio ns,

    T. Mekki, I. Jabri, A. Rachedi, and M. ben Jemaa, “V ehicu lar cloud networks: Challenges, architectures, and future directio ns,” V ehicular Communications, vol. 9, pp. 268–280, 2017

    work page 2017

  11. [11]

    Imp roving aodv with preemptive local route repair,

    S. Crisstomo, S. Sargento, P . Brando, and R. Prior, “Imp roving aodv with preemptive local route repair,” Technical Report Series:DCC-2003- 0, Universidade do Porto , 2003

    work page 2003

  12. [12]

    A performance evaluation of a preemptive on-demand distance vector routing protocol for mobile adho c networks,

    A. Boukerche and L. Zhang, “A performance evaluation of a preemptive on-demand distance vector routing protocol for mobile adho c networks,” wireless communications and mobile computing , vol. 4, p. 99108, 2004

    work page 2004

  13. [13]

    Predictive preemptive adhoc on demand distance vector routing,

    S. B. Hacene, A. Lehireche, and A. Meddahi, “Predictive preemptive adhoc on demand distance vector routing,” Malaysian Journal of Com- puter Science , vol. 19, no. 2, pp. 189–195, 2006

    work page 2006

  14. [14]

    High p recision- predictive preemptive adhoc on demand distance vector rout ing in ad hoc networks,

    H. Kitamoto, N. Masuda, S. H. Bouk, and I. Sasase, “High p recision- predictive preemptive adhoc on demand distance vector rout ing in ad hoc networks,” International Symposium on Wireless Personal Multimedia Communications, 2008

    work page 2008

  15. [15]

    Prediction based routing for vehicular ad hoc networks,

    V . Namboodiri and L. Gao, “Prediction based routing for vehicular ad hoc networks,” IEEE Transactions on V ehicular Technology, vol. 56, pp. 2332–2345, July 2007

    work page 2007

  16. [16]

    Probability distribution of lin k life time in vehicular ad hoc networks,

    S. Shelly and A. Babu, “Probability distribution of lin k life time in vehicular ad hoc networks,” International Conference on Advances in Computing, Communications and Informatics (ICACCI) , pp. 1418–1423, 2013

    work page 2013

  17. [17]

    Performance evaluat ion of link metrics in vehicle networks: A study from the cologne ca se,

    J. Zhang, M. Y . Ren, and H. Labiod, “Performance evaluat ion of link metrics in vehicle networks: A study from the cologne ca se,” ACM Symposium on Development and Analysis of Intelligent V e hicular Networks and Applications , pp. 123–130, 2016

    work page 2016

  18. [18]

    Advance a nd prospects of adaboost algorithm,

    C. Ying, M. Qi-Guang, L. Jia-Chen, and G. Lin, “Advance a nd prospects of adaboost algorithm,” Acta Automatica Sinica , vol. 39, no. 6, pp. 745– 758, 2013

    work page 2013

  19. [19]

    Eldp: ext ended link duration prediction model for vehicular networks,

    X. Wang, C. Wang, G. Cui, Q. Y ang, and X. Zhang, “Eldp: ext ended link duration prediction model for vehicular networks,” International Journal of Distributed Sensor Networks , vol. 12, no. 4, p. 5767569, 2016

    work page 2016

  20. [20]

    A new cooperative communication algorithm for improving connectivity in the event of network failure i n vanets,

    B. Das and J. Almhana, “A new cooperative communication algorithm for improving connectivity in the event of network failure i n vanets,” Computer Networks , vol. 128, pp. 51–62, 2017

    work page 2017

  21. [21]

    Predicting link f ailure in vehicular communication system using link existence diagr am (led),

    S. K. Bhoi, M. Singh, and P . M. Khilar, “Predicting link f ailure in vehicular communication system using link existence diagr am (led),” in Progress in Intelligent Computing Techniques: Theory, Pra ctice, and Applications. Springer, 2018, pp. 501–506

    work page 2018

  22. [22]

    Support vector r egression,

    D. Basak, S. Pal, and D. C. Patranabis, “Support vector r egression,” Neural Information Processing-Letters and Reviews , vol. 11, no. 10, pp. 203–224, 2007

    work page 2007