Pith. sign in

REVIEW 4 major objections 5 minor 36 references

E-CONDOR: Efficient Contour-Based Detection Of Random Spatial Signals From UAV Observations Using Dual Stochastic Gradient

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Dual stochastic gradient maps fields with less drone flying distance for a given map accuracy.

desk verdict A legible, incremental extension of the authors' earlier WSN contour work to UAVs, but the dual-SG efficiency claim is not yet proven: the control error (Eq. 3) and the evaluation error (Eq. 10) are different quantities, and the evidence is a single synthetic field without error bars. read the letter →

arxiv 2411.17449 v1 pith:RYZT7KKZ submitted 2024-11-26 eess.SP

Maryam Zahra , Homa Tajiani , Hadi Alasti This is my paper · ORCID
classification eess.SP
keywords UAVspatialmonitoringcontourlinetracingdualstochasticgradientdatafusioncentersignalmodelingmeanabsoluteerrorLloyd-Maxquantizationefficiency
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 claims that a drone can map a spatially correlated field by tracing iso-contour lines at a few chosen levels, and that two stochastic-gradient routines at the data fusion center make this tracing schedule self-improving. The first gradient decides how many new contour levels to add in each round; the second drops any new level that falls too close to an already-traced contour. In simulations, this dual-gradient schedule converges faster, reaches a lower mean absolute modeling error, and needs less total flying distance for a given target accuracy than the same contour-tracing scheme without the gradients. The practical payoff is cheaper environmental monitoring because the field is represented by contour traces rather than dense grid samples.

What carries the argument

The load-bearing mechanism is the dual stochastic-gradient pair acting on the contour-level schedule rather than on the field estimate directly. The data fusion center updates the level-count increment $$\kappa_n = \kappa_{n-1} + \left\lceil 1 + \frac{2|\mathrm{Error}_{n-2} - \mathrm{Error}_{n-1}|}{|\mathrm{Error}_{n-2} + \mathrm{Error}_{n-1}|} \right\rceil$$ and the redundancy radius $$\delta_n = \delta_{n-1}\left|1 - \frac{2(\mathrm{Error}_{n-2} - \mathrm{Error}_{n-1})}{\mathrm{Error}_{n-2} + \mathrm{Error}_{n-1}}\right|$$ using the learning error of the latest spline reconstruction. Because the levels themselves are Lloyd-Max-spaced and redundant levels are pruned before they are sent to the UAVs, the loop acts as a self-tightening contour sampler.

What would settle it

Re-run the paper's simulation with zero-mean Gaussian noise added to the UAV-reported coordinates and compare the dual-SG and no-SG MAE-versus-distance curves; if the dual-SG curve no longer stays below the no-SG curve at low MAE, the claimed efficiency gain depends on the noise-free assumption.

Watch

Extended reading notes

Core claim

The central discovery is that the contour-level schedule itself can be treated as a learnable parameter. After each round of UAV tracing, the data fusion center reconstructs the field with a bipolar spline interpolator, computes a learning error, and uses that error in two stochastic-gradient updates: $\kappa_n$ grows the number of levels to trace, while $\delta_n$ shrinks the neighborhood that marks redundant levels. New levels are Lloyd-Max-optimal for the current field estimate, so each round samples the signal where it carries the most new information. The paper's simulations show that this dual-SG schedule reaches a target spatial modeling error with less total trace length than the same contour-tracing procedure without the gradient updates.

Load-bearing premise

The load-bearing premise is that the coordinates a UAV reports lie exactly on the true contour at the assigned level, with no observation noise and no drift in the field while the contours are being traced.

Editorial extensions

If this is right

  • The same target MAE is reached with less total flying distance, and the largest savings appear at the low-MAE end where mapping is most demanding.
  • The number of contour levels needed is determined automatically from the learning error, so the operator does not have to fix the level count in advance.
  • Redundant contour levels are dropped before any UAV is sent to trace them, cutting both transmitted coordinates and mission time.
  • Convergence in the sense of the learning error takes fewer iterations, so the field model stabilizes in fewer rounds of UAV sorties.

Reading between the lines

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

  • Adding realistic GPS and sensor noise to the reported coordinates would perturb the on-contour assumption in Eq. (10), so the efficiency gap shown in Fig. 7 could shrink; testing this would establish how much of the gain survives real hardware.
  • The dual-SG scheduler could be repurposed for multi-UAV fleets by interpreting the batch size as the number of contours traced in parallel, turning the redundancy radius into a fleet-coordination parameter.
  • If the field's signal-strength distribution is not known ahead of the first pass, the Lloyd-Max level placement would have to be learned from the spline estimate, coupling the level spacing and the gradient updates into one online estimation problem.
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 manuscript proposes E-CONDOR, an algorithm for UAV-based spatial monitoring in which a data fusion center selects iso-contour levels for UAVs to trace, reconstructs the field using a spline interpolator, and adaptively decides both how many new contour levels to add and which previously traced levels are redundant. The adaptive decisions are made by two updates, Eqs. (6) and (7), that the paper calls dual stochastic gradient rules. The paper reports simulation results on a single synthetic correlated field, claiming that the dual-SG version converges faster, achieves lower spatial modeling MAE, and requires less total UAV flying distance for a given MAE than a baseline without SG.

Significance. If the claimed efficiency gains are substantiated, the algorithm would be a practically useful contribution to UAV-based environmental monitoring, since it directly targets the dominant cost, which is UAV flying distance. The paper has clear strengths: the algorithm is described as an explicit procedure, the synthetic field model is specified, and the authors provide a link to MATLAB code for verification. The evaluation metric, true MAE against the actual field (Eq. 10), is an external benchmark rather than a purely self-referential one. However, the evidence presented is currently too thin to support the headline claims: the core updates are introduced without derivation or convergence analysis, the baseline is not precisely defined, and all conclusions rest on a single synthetic field realization with no statistical characterization.

major comments (4)
  1. [Section IV-B, Eqs. (6) and (7)] The two updates are called stochastic gradient rules, but no objective function, loss gradient, or convergence argument is given. The reader cannot verify that these equations are gradients of any quantity, nor that iterating them drives either the learning error or the true MAE to a minimum. Because these equations are the central novelty of the paper, the authors should either derive them from a well-defined objective (e.g., minimizing expected MAE as a function of κ and δ) or, failing that, clearly present them as heuristic adaptive schedules and support them with a convergence analysis or a much more extensive empirical study.
  2. [Section V-B, Eq. (3) versus Eq. (10)] The controller updates κ and δ using Error_n from Eq. (3), which measures the mean absolute difference between successive reconstructions g̃_n and g̃_{n-1}. The reported performance, however, is E_n from Eq. (10), the mean absolute error against the true field g. A reconstruction can stop changing while remaining substantially biased, so a small Error_n does not imply a small E_n. The paper provides no derivation or simulation evidence that driving Error_n down drives E_n down. This is a load-bearing gap because the headline efficiency claim in Fig. 7 is stated in terms of E_n per unit flying distance. At minimum, the authors should plot both Error_n and E_n on the same runs and show that they track each other, or provide a bound relating the two.
  3. [Section V-B, Figs. 3, 6, and 7] The comparison baseline, labeled 'without dual SG,' is not specified precisely. It is not stated whether the baseline uses the same Lloyd-Max level computation, the same bipolar spline interpolator, the same initialization procedure, the same convergence criterion, or a fixed increment schedule for the number of levels, nor whether redundant-level dropping is disabled entirely. Without a precise baseline algorithm, the advantage shown in Figs. 3 and 7 could be an artifact of a weaker baseline rather than of the dual SG updates themselves. The authors should provide a step-by-step description of the baseline and ensure that the only difference is the presence of the SG updates.
  4. [Section V-A and V-B] All reported curves are obtained from a single synthetic field realization (Fig. 2), with no error bars, no Monte Carlo runs over random fields, and no sensitivity analysis with respect to κ_0, δ_0, the convergence threshold, the Gaussian parameters σ1 and σ2, or the number of Gaussian components. Since the paper claims 'significantly' faster convergence and lower cost, these claims need statistical support. The authors should repeat the experiments over many random fields and report means and confidence intervals, and also test a range of algorithm parameters.
minor comments (5)
  1. [Section III, Eq. (2) and surrounding text] The notation for Error is inconsistent: the text defines ∇Error_{n-1} = Error_{n-1} - Error_{n-2} and then writes Error_{n-1} = 1/2(Error_{n-1} + Error_{n-2}), which is circular and likely a typographical error. Please introduce distinct symbols for the mean error and the error difference.
  2. [Summary of the algorithm, step 8] Step 8 says 'n + 1 ← n'; this should likely be 'n ← n + 1' or an equivalent increment statement.
  3. [Section V-B-2] There is a typo in the text: 'The signal span range is defined as: as: sprn = SPn / SPActual.' The duplicate 'as:' should be removed.
  4. [Section V-B-1, Eq. (11)] The definition MAE_n (dB) = 20 log10(E_n) should specify whether this is a power-like or amplitude-like conversion, and why 20 is chosen rather than 10. Also, 'Log10' should be typeset as 'log10'.
  5. [Section IV-B] The paper does not explain how the UAV finds the 'actual coordinate of the initiation coordinate' when the assigned point lies near a contour but the field estimate is rough. A short description of the search procedure would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the headline MAE comparison is measured against synthetic ground truth, not against the controller's self-referential learning-error sequence.

full rationale

The paper's central efficiency claim is evaluated externally. The dual-SG updates in Eqs. (6) and (7) use the model-change error Error_n of Eq. (3) as a feedback signal for choosing kappa and delta; no parameter is fitted to minimize the true spatial modeling MAE of Eq. (10). The headline plots, Fig. 3 and Fig. 7, report E_n = (1/PQ) sum |g - g_tilde_n| against the synthetic ground truth field of Eqs. (8)-(9), which is independent of the controller's update rule and of the stopping heuristic based on a 'tangible drop' in the learning error. The numerous self-citations [1]-[5], [23]-[24] supply the background contour/SG framework, but the dual-SG algorithm is implemented and evaluated in this paper, and the code is available online [36]; no load-bearing mathematical result is imported by authority. The acknowledged noiseless and time-invariant assumptions, and the underspecified 'without SG' baseline, are validity or reproducibility concerns rather than circularity. No equation in the derivation is equivalent to its inputs by construction, and no fitted quantity is renamed as a prediction. Therefore no significant circularity is found.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The algorithm rests on several domain assumptions about noise-free, time-invariant contour tracing and on the unexplained heuristic updates in Eqs. (6) and (7). It introduces no new physical entities; the free parameters are the initial values of the two adaptive factors and the convergence threshold, none of which are specified in the paper.

free parameters (3)
  • κ_0 (initial level-increment factor) = not specified
    The update (6) requires a starting κ_{n-1}; no initial value or tuning procedure is given in the paper.
  • δ_0 (initial redundancy margin) = not specified
    The update (7) requires a starting δ_{n-1}; the paper does not specify how this is initialized or normalized to the signal range.
  • convergence threshold = not specified
    The stopping condition 'learning error is tangibly dropped' is qualitative; no numerical threshold is given.
assumptions (5)
  • domain assumption The spatial signal remains unchanged during the multi-contour observation period.
    Section II states this assumption, which is required for the reported coordinates from different contour levels to describe one consistent field.
  • domain assumption UAV observations are noise-free (negligible noise).
    Section I and V-A ignore observation noise, so the reported trace coordinates are treated as exactly lying on the iso-contour.
  • domain assumption Lloyd-Max quantizer produces optimally spaced contour levels for the current estimate.
    Section IV-B uses Eqs. (4) and (5) assuming the estimated PDF yields levels that minimize modeling error, which presumes the estimate's PDF approximates the true field.
  • domain assumption Bipolar spline interpolation reconstructs a faithful field estimate from sparse contour coordinates.
    Section IV-B uses the algorithm in [32] without discussing its error or applicability to non-smooth fields.
  • standard math The error sequence Error_n is strictly positive so divisions in Eqs. (6) and (7) are well-defined.
    The absolute values in (7) mitigate sign issues, but (6) divides by |Error_{n-2} + Error_{n-1}|; a zero error would break the update.

how reviews work

0 comments
Cite this review

Pith. "Pith review of E-CONDOR: Efficient Contour-Based Detection Of Random Spatial Signals From UAV Observations Using Dual Stochastic Gradient." pith.science (2026). https://pith.science/paper/RYZT7KKZ

@misc{pith2026241117449,
  author       = {Pith},
  title        = {Pith review of: E-CONDOR: Efficient Contour-Based Detection Of Random Spatial Signals From UAV Observations Using Dual Stochastic Gradient},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYZT7KKZ}},
  note         = {Machine review of arXiv:2411.17449}
}
read the original abstract

This paper presents a novel efficient method for spatial monitoring of the distribution of correlated field signals, such as temperature, humidity, etc. using unmanned aerial vehicles (UAVs). The spatial signal is compressed to its iso-contour lines at a number of known levels that are introduced by data fusion center (DFC). The UAV traces a contour line of the field signal at a time, and reports the coordinates of its own traces to the DFC for spatial modeling. The DFC iteratively improves the spatial model of the field signal and assigns a new contour level to each UAV to trace and report its coordinates for spatial model improvement. The selected batch of levels and the start point of the search are introduced by the DFC. In order to reduce the required data for spatial modeling, and accordingly improve the algorithm data efficiency, dual stochastic gradient routines are used at the DFC to find a next proper number of contour levels in the batch, and to eliminate the redundant contour levels, in each iteration. The performance evaluation of the proposed algorithm based on computer simulations demonstrates significantly faster convergence, better signal estimation, and a higher data efficiency against when the stochastic gradient is not used.

Figures

Figures reproduced from arXiv: 2411.17449 by the authors.

Figure 1
Figure 1. The spatial distribution is compressed into its contour lines, and a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The original spatial signal: (a) is estimated based on a first rough [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. The spatial learning error of Fig. 2 versus the number of levels. The [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (3 more)
Figure 3
Figure 3. Figure 3: The spatial modeling error of Fig. 2 versus the number of levels. The [PITH_FULL_IMAGE:figures/full_fig_p004_3.png]
Figure 6
Figure 6. Figure 6: The figure compares the total flying distance of the UAV (the cost), [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: The spatial monitoring performance (MAE in dB), versus the total [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 28 canonical work pages

  1. [1]

    Alasti, ”Efficient Sensing of Correlated Spatiotemporal Signals: A Stochastic Gradient Approach,” IEEE Signal Processing Letters, 2019

    H. Alasti, ”Efficient Sensing of Correlated Spatiotemporal Signals: A Stochastic Gradient Approach,” IEEE Signal Processing Letters, 2019

    work page 2019

  2. [2]

    Alasti, ”Efficient Tracking of Spatially Correlated Signals in Wireless Sensor Fields: A Weighted Stochastic Gradient Approach,” V ol

    H. Alasti, ”Efficient Tracking of Spatially Correlated Signals in Wireless Sensor Fields: A Weighted Stochastic Gradient Approach,” V ol. 11, no. 2, pp. 78-90, 2021, https://doi.org/10.1049/wss2.12012

    work page doi:10.1049/wss2.12012 2021

  3. [3]

    H. Alasti, ”An Efficient Machine Learning Algorithm For Spatial Tracking Of Correlated Signals In Wireless Sensor Field,” 2020 30th International Telecommunication Networks and Applications Con- ference (ITNAC), Melbourne, VIC, Australia, 2020, pp. 1-6, doi: 10.1109/ITNAC50341.2020.9315016

    work page arXiv 2020

  4. [4]

    H. Alasti, ”An Efficient Accelerated Learning Algorithm For Tracking Of Unknown, Spatially Correlated Signals In Ad-Hoc Wireless Sensor Networks,” 2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON), New York, NY , USA, 2020, pp. 0813-0819, doi: 10.1109/UEMCON51285.2020.9298182

    work page arXiv 2020

  5. [5]

    Alasti, ”Communication-Efficient Tracking of Unknown, Spatially Correlated Signals in Ad-Hoc Wireless Sensor Networks: Two Ma- chine Learning Approaches,” Sensors, vol

    H. Alasti, ”Communication-Efficient Tracking of Unknown, Spatially Correlated Signals in Ad-Hoc Wireless Sensor Networks: Two Ma- chine Learning Approaches,” Sensors, vol. 21, n. 15, pp. 1-18, 2021, . https://doi.org/10.3390/s21155175

    work page doi:10.3390/s21155175 2021

  6. [6]

    Seo and H

    J. Seo and H. Park, ”Object Recognition in Very Low Resolution Images Using Deep Collaborative Learning,” in IEEE Access, vol. 7, pp. 134071- 134082, 2019, doi: 10.1109/ACCESS.2019.2941005

    arXiv 2019

  7. [7]

    L. Gao, R. Zhang, L. Qi, E. Chen and L. Guan, ”The Labeled Mul- tiple Canonical Correlation Analysis for Information Fusion,” in IEEE Transactions on Multimedia, vol. 21, no. 2, pp. 375-387, Feb. 2019, doi: 10.1109/TMM.2018.2859590

    work page arXiv 2019

  8. [8]

    L. Tang, Z. Yang and K. Jia, ”Canonical Correlation Analysis Regu- larization: An Effective Deep Multiview Learning Baseline for RGB- D Object Recognition,” in IEEE Transactions on Cognitive and De- velopmental Systems, vol. 11, no. 1, pp. 107-118, March 2019, doi: 10.1109/TCDS.2018.2866587

    arXiv 2019

Show all 36 references
  1. [9]

    Zhang et al., ”Detecting Object Open Angle and Direction Using Machine Learning,” in IEEE Access, vol

    Y . Zhang et al., ”Detecting Object Open Angle and Direction Using Machine Learning,” in IEEE Access, vol. 8, pp. 12300-12306, 2020, doi: 10.1109/ACCESS.2020.2965537

    2020

  2. [10]

    Hussein, P

    S. Hussein, P. Kandel, C. W. Bolan, M. B. Wallace and U. Bagci, ”Lung and Pancreatic Tumor Characterization in the Deep Learning Era: Novel Supervised and Unsupervised Learning Approaches,” in IEEE Transactions on Medical Imaging, vol. 38, no. 8, pp. 1777-1787, Aug. 2019, doi:...

    2019

  3. [11]

    S. Lv, Y . Lu, M. Dong, X. Wang, Y . Dou and W. Zhuang, ”Qualitative Action Recognition by Wireless Radio Signals in Human–Machine Sys- tems,” in IEEE Transactions on Human-Machine Systems, vol. 47, no. 6, pp. 789-800, Dec. 2017, doi: 10.1109/THMS.2017.2693242

    2017

  4. [12]

    Chaki, A

    S. Chaki, A. Routray and W. K. Mohanty, ”Well-Log and Seismic Data Integration for Reservoir Characterization: A Signal Processing and Machine-Learning Perspective,” in IEEE Signal Processing Magazine, vol. 35, no. 2, pp. 72-81, March 2018, doi: 10.1109/MSP.2017.2776602

    2018

  5. [13]

    A. M. Alzahed, S. Mikki and Y . Antar, ”Electromagnetic Machine Learning for Inverse Modeling Using the Spatial Singularity Expansion Method,” in IEEE Journal on Multiscale and Multiphysics Computational Techniques, vol. 5, pp. 59-71, 2020, doi: 10.1109/JMMCT.2020.2973101

    2020

  6. [14]

    Wickramasinghe, D

    A. Wickramasinghe, D. C. Ranasinghe, C. Fumeaux, K. D. Hill and R. Visvanathan, ”Sequence Learning with Passive RFID Sensors for Real-Time Bed-Egress Recognition in Older People,” in IEEE Journal of Biomedical and Health Informatics, vol. 21, no. 4, pp. 917-929, July 2017, doi...

    2017

  7. [15]

    J. I. Hamilton and N. Seiberlich, ”Machine Learning for Rapid Mag- netic Resonance Fingerprinting Tissue Property Quantification,” in Pro- ceedings of the IEEE, vol. 108, no. 1, pp. 69-85, Jan. 2020, doi: 10.1109/JPROC.2019.2936998

    2020

  8. [16]

    Asokan and J

    A. Asokan and J. Anitha, ”Machine Learning based Image Processing Techniques for Satellite Image Analysis -A Survey,” 2019 Interna- tional Conference on Machine Learning, Big Data, Cloud and Parallel Computing (COMITCon), Faridabad, India, 2019, pp. 119-124, doi: 10.1109/COMIT...

    2019

  9. [17]

    A. Correa Silva, ”Meta-Learning Applications in Digital Image Pro- cessing,” 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), Niter´oi, Brazil, 2020, pp. 19-20, doi: 10.1109/IWS- SIP48289.2020.9145478

    2020

  10. [18]

    M. Adel, A. Kotb, O. Farag, M. S. Darweesh and H. Mostafa, ”Breast Cancer Diagnosis Using Image Processing and Machine Learning for Elastography Images,” 2019 8th International Conference on Modern Circuits and Systems Technologies (MOCAST), Thessaloniki, Greece, 2019, pp. 1-4...

    2019

  11. [19]

    L. Sun, W. Shao, M. Wang, D. Zhang and M. Liu, ”High-Order Feature Learning for Multi-Atlas Based Label Fusion: Application to Brain Segmentation With MRI,” in IEEE Transactions on Image Processing, vol. 29, pp. 2702-2713, 2020, doi: 10.1109/TIP.2019.2952079

    2020

  12. [20]

    Sivanathan et al., ”Classifying IoT Devices in Smart Environments Using Network Traffic Characteristics,” in IEEE Transactions on Mo- bile Computing, vol

    A. Sivanathan et al., ”Classifying IoT Devices in Smart Environments Using Network Traffic Characteristics,” in IEEE Transactions on Mo- bile Computing, vol. 18, no. 8, pp. 1745-1759, 1 Aug. 2019, doi: 10.1109/TMC.2018.2866249

    2019

  13. [21]

    Alasti, Non-uniform-level crossing sampling for efficient sensing of temporally sparse signals, IET Wireless sensor systems, vol

    H. Alasti, Non-uniform-level crossing sampling for efficient sensing of temporally sparse signals, IET Wireless sensor systems, vol. 4, no. 1, pp. 27-34, 2014

    2014

  14. [22]

    H. Alasti, Level crossing sampling for energy conservation in wireless sensor networks: A design framework, , book chapter in Wireless Tech- nologies: Concepts, Methodologies, Tools and Applications, IGI Global, 2011

    2011

  15. [23]

    Alasti, ”An on-demand compressed sensing approach for spatial monitoring of correlated big data using multi-contours in dense wireless sensor network” Proc

    H. Alasti, ”An on-demand compressed sensing approach for spatial monitoring of correlated big data using multi-contours in dense wireless sensor network” Proc. IEEE Intl. Conf. Wireless for Space and Extreme Environments (WiSEE), Montreal, Canada, pp 86-91, Oct. 2017

    2017

  16. [24]

    Alasti, A

    H. Alasti, A. Nasipuri, Spatiotemporal monitoring using contours in Large-scale wireless sensor networks, ACM FOW ANC09, May 18, pp. 77-85, 2009, New Orleans, Louisiana, USA

    2009

  17. [25]

    Alasti, Energy efficient spatiotemporal threshold level detection in large scale wireless sensor fields, IEEE Online Conference on Green Communications (GreenCom), pp

    H. Alasti, Energy efficient spatiotemporal threshold level detection in large scale wireless sensor fields, IEEE Online Conference on Green Communications (GreenCom), pp. 151-156, 2012

    2012

  18. [26]

    Alasti, W.A

    H. Alasti, W.A. Armstrong, A. Nasipuri, Performance of a robust filter- based approach for contour detection in wireless sensor networks, in Proceedings of ICCCN’07, pages 159-164, 2007

    2007

  19. [27]

    Buragohain, S

    C. Buragohain, S. Gandhi, J. Hershberger, S. Suri, Contour approxima- tion in sensor networks, DCOSS ’06, June 18-20, 2006

    2006

  20. [28]

    Gandhi, J

    S. Gandhi, J. Hershberger, S. Suri, Approximate iso-contours and spatial summaries in sensor networks, IPSN’07, April 25-27, 2007

    2007

  21. [29]

    Sarkar, X

    R. Sarkar, X. Zhu, J. Gao, L.J. Guibas, J.S.B. Mitchell, Iso-contour queries and gradient descent with guaranteed delivery in sensor networks, in Proceedings of IEEE INFOCOM’08, 2008

    2008

  22. [30]

    J. Lian, L. Chen, K. Naik, Y . Liu, G.B. Agnew, Gradient boundary detection for time series snapshot construction in sensor networks, IEEE Transactions on Parallel and Distributed Systems, vol.18, no.10, pp. 1462- 1475, Oct. 2007

    2007

  23. [31]

    Bershad, Analysis of the normalized LMS algorithm with Gaussian inputs, in IEEE Transactions on Acoustics, Speech, and Signal Processing, p

    N. Bershad, Analysis of the normalized LMS algorithm with Gaussian inputs, in IEEE Transactions on Acoustics, Speech, and Signal Processing, p. 793-706, vol. 34, no. 4,1986

    1986

  24. [32]

    Sandwell, Bipolar spline interpolation of GEOS-3 and SEASAT altimeter data, Geophysical Research Letters, vol

    D.T. Sandwell, Bipolar spline interpolation of GEOS-3 and SEASAT altimeter data, Geophysical Research Letters, vol. 14, no. 3, pp. 139- 142, Feb. 1987

    1987

  25. [33]

    Sayood, Introduction to data compression, Published by Morgan Kaufmann, 2000

    K. Sayood, Introduction to data compression, Published by Morgan Kaufmann, 2000

    2000

  26. [34]

    H. Alasti, ”On Norm Selection Effect In Energy Efficient Modeling of Correlated Spatial Signals Using Machine Learning in Wireless Sensor Field,” 2021 International Conference on Green Energy, Com- puting and Sustainable Technology (GECOST), 2021, pp. 1-5, doi: 10.1109/GECOST5...

    2021

  27. [35]

    Jindal, K

    A. Jindal, K. Psounis, Modeling spatially correlated data in sensor networks, ACM Transactions on Sensor Networks, November 2006

    2006

  28. [36]

    Alasti, (2024) Efficient spatiotemporal monitoring of corre- lated signals based on dual stochastic gradient [Online]

    H. Alasti, (2024) Efficient spatiotemporal monitoring of corre- lated signals based on dual stochastic gradient [Online]. Available: https://github.com/HarryAlasti/UA V-1-codes-2024

    2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.