Pith. sign in

REVIEW 4 major objections 4 minor 25 references

SMDS-based Rigid Body Localization

T0 review · 4 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper proposes an SMDS variant that recovers a rigid body's translation and rotation from distance and angle measurements using a sub-block of the complex edge kernel, with simulated MSE close to the CRLB.

desk verdict A useful SMDS-to-RBL extension whose central update equation depends on the very unknown edges it claims to estimate; worth a conditional review but needs a rewrite. read the letter →

arxiv 2509.01223 v1 pith:PBB3QP7D submitted 2025-09-01 eess.SP

Niclas Führling , Giuseppe Abreu , David González G. , Osvaldo Gonsa This is my paper · ORCID
classification eess.SP
keywords rigidbodylocalizationsupermultidimensionalscalingcomplexedgekernelangleofarrivaldistancemeasurementsCramér-Raolowerboundwirelesssensornetworks
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 tries to establish that a moving rigid body can be localized from mixed distance and angle measurements using only a small part of the complex edge kernel of the SMDS algorithm. If true, the position, shape, and orientation of the body can be recovered with less computational and measurement overhead than constructing and truncating the full kernel. The paper's simulations compare the translation and rotation estimates against a prior MDS-based method and against the CRLB, and report that the estimates reach the fundamental limit. The reason this matters is that rigid body localization underlies navigation, collision detection, and vehicle path prediction, where knowing orientation as well as position is necessary.

What carries the argument

The central object is the rank-one complex edge kernel K = v^* v^T, whose entries are products of directed complex edges between pairs of nodes. The method uses a three-block minor of K that couples the unknown anchor-to-target edge vector v_AT with the known anchor-to-anchor and target-to-target edge vectors; the fixed-point update of Eq. (14) is the mechanism that recovers v_AT from that minor. The rank-one structure ensures the kernel contains a consistent geometry, while the minor selection is what makes the method cheaper than full SMDS and tailored to the rigid-body measurement setup.

What would settle it

In an 8-anchor, 8-target simulation, withhold all anchor-to-target distance and angle data, form K3 using only products of anchor-to-anchor and target-to-target edges, and run Eq. (14). If the recovered v_AT does not approach the CRLB in MSE, or if K3 is not uniquely determined by the known blocks, the paper's information premise is false.

Watch

Extended reading notes

Core claim

The central claim is that a rigid body's landmark positions, translation, and rotation can be recovered by solving for the anchor-to-target edge vector v_AT instead of the full edge vector v. The complete SMDS kernel K = v^* v^T is rank one and would normally be truncated by its largest eigenpair; this paper instead selects the three blocks K1, K3, K4 that couple v_AT with the known anchor-to-anchor and target-to-target edges, and turns them into the fixed-point update of Eq. (14). Numerical simulations with 8 anchors and an 8-point rigid body in a 10 m-by-10 m room are intended to show that the resulting estimate of translation and rotation is close to the CRLB and better than a distance-on

Load-bearing premise

The load-bearing premise is that the sub-block of the kernel needed to update the anchor-to-target edges can be assembled from anchor-to-anchor and target-to-target measurements alone; if that sub-block actually requires the unknown anchor-to-target measurements themselves, the iteration is circular.

Editorial extensions

If this is right

  • If the central claim holds, an SMDS-based estimator can output the rigid body's translation and rotation at CRLB-level accuracy while only forming a sub-block of the edge kernel, reducing complexity compared with full SMDS.
  • The method offers a unified treatment of distance and angle data: when angle measurements are unavailable, a preliminary MDS step supplies angles, and the same kernel update then applies.
  • The iterative structure of Eq. (14) means the method can be initialized from anchor-to-anchor and target-to-target data alone and refined as anchor-to-target information enters.
  • The comparison against the CRLB positions the method as practically efficient, since its MSE tracks the theoretical lower bound over the simulated range-error interval.

Reading between the lines

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

  • The authors leave implicit that Eq. (14) needs K3 = v_AT^* v_AT^T, a block quadratic in the very anchor-to-target edges being estimated. If K3 is built from noisy anchor-to-target measurements, part of the reported accuracy may come from those measurements directly rather than from the SMDS minor structure.
  • A natural ablation would compare the full method, the distance-only variant, and a direct least-squares estimator that uses anchor-to-anchor and target-to-target information without the kernel minor; this would isolate what the SMDS update contributes.
  • The same sub-block construction should carry over to 3D with a suitable quaternion or rotation-constrained representation, since the rank-one kernel identity itself is dimension-independent.
  • In settings with sparse target-to-target measurements, the method's dependence on the known v_TT block would make the update degrade; testing the iteration against the fraction of missing target-to-target edges would map its operating range.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes a rigid body localization (RBL) method based on an SMDS-style rank-one complex edge kernel. The method is intended to estimate the anchor-to-target edge vector v_AT, and then the target positions, using only a sub-block of the kernel (Eq. (13)). From v_AT and the known rigid-body shape, translation and rotation are recovered via a standard weighted least-squares fit (Section III.B). Simulation results compare translation and rotation MSE against CRLBs for distance-only and distance-plus-angle variants. The central claim is that the sub-block depends on anchor-to-anchor and target-to-target information only, and that the resulting estimates approach the CRLB.

Significance. If the method worked as claimed, it would be a useful low-complexity alternative to full SMDS for RBL: it would exploit the known structure of the anchors and the rigid body while estimating only the anchor-to-target block. The rank-one kernel construction is correct in the noiseless model, and the rotation/translation least-squares step is a standard and sound component. The paper also provides a comparison against CRLBs from a companion paper, which is appropriate if the measurement model matches. However, the central algorithmic claim is not currently supported: the selected sub-block is not constructed from AA/TT information alone, and Eq. (12) and Eq. (14) contain serious technical gaps. The contribution therefore needs substantial revision before it can be evaluated as stated.

major comments (4)
  1. [Section III.A, Eqs. (13)-(15), and Conclusion] The paper's key claim is that the used minor of the kernel depends only on anchor-to-anchor and target-to-target information. This is not correct. The block column in Eq. (13) contains K1 = v_AA^* v_AT^T, K3 = v_AT^* v_AT^T, and K4^T = v_TT^* v_AT^T. Every one of these blocks is quadratic in the unknown v_AT. If only AA and TT edges are known, K1, K3, and K4 cannot be computed. The initial estimate in Eq. (15) also requires K1 and K4. Thus the algorithm, as described, cannot be executed under the information model stated in the Abstract and Conclusion. The authors must either state explicitly that noisy AT measurements are used to form these blocks, or show how they are obtained from AA/TT information alone. If K3 is formed from the current iterate, the noiseless update becomes circular and the claimed 'key point' must be revised.
  2. [Section II.B, Eq. (12)] Eq. (12) writes x_hat = C^{-1} v_hat, but C in Eq. (6) is rectangular: v has P = T(T-1)/2 entries while x has T entries. Even disregarding the rank deficiency from the translation/rotation gauge, C^{-1} does not exist. A pseudo-inverse or a constrained least-squares inversion with anchor coordinates must be specified. The reader cannot reproduce the reconstruction step from the text as written. Citing [21] is not sufficient, because this paper's v_AT-only formulation changes the problem.
  3. [Section III.A, Eq. (14)] Eq. (14) is asserted without derivation. The notation is ambiguous: the numerator appears to be a row-vector product, the denominator is unclear, and no convergence proof is given. The noiseless fixed point may reproduce v_AT by construction if K3 is built from v_AT, but this does not establish convergence from a noisy initial estimate or uniqueness. A rigorous derivation of the update, an explicit statement of which quantities are known, and a convergence analysis are required before the method can be assessed.
  4. [Section IV.B, Figs. 2-3] The simulations do not describe how the kernel sub-blocks are formed in practice. In particular, it is not stated whether K1, K3, and K4 are computed from noisy AT measurements, from true v_AT, or from the current iterates. This makes it impossible to distinguish a genuine estimator from one that uses oracle knowledge of the unknown edges. Please provide a complete algorithmic specification, including how each block of Eq. (13) is obtained from the measurements, and ideally release code or pseudocode.
minor comments (4)
  1. [Abstract/Conclusion] There are grammatical errors, e.g., 'an dthe' in the Conclusion. More importantly, the wording 'based only on a set of measurements of the distances, as well as the angles between sensors of the vehicle to the anchor landmark points' seems to describe AT measurements, while the subsequent 'anchor-to-anchor and target-to-target information' describes known geometry. The information model should be stated unambiguously.
  2. [Section II.B, Eq. (6)] The symbol M is used both for the number of anchors and for the total number of edges. This is confusing; use a distinct symbol such as P for the edge count, as introduced in the text.
  3. [Section IV.B] Figure 2 is labeled RMSE while Eq. (23) and the text define MSE. Also the text says that the distance-only SMDS performs slightly better than full SMDS in the small-range-error regime, but this is not explained; if SMDS is not optimal in that regime, a comment is needed.
  4. [Section IV.A, Eqs. (24)-(25)] The notation g'|^d_t and lambda_na is not defined in enough detail to reproduce the CRLB. Since the CRLB is taken from the companion paper [25], either summarize the relevant definitions or include them in an appendix.

Circularity Check

1 steps flagged · score 6.0 of 10

Self-definitional AT block: Eq. (14) rearranges K3 = v_AT^*v_AT^T, so the estimated edge vector is an input to the algorithm; the advertised AA/TT-only minor is unobservable.

  1. self definitional [Eq. (13)-(14), Section III.A; Abstract/Conclusion]
    "which yields [K1; K3; K4^T] = [v_AA^*; v_AT^*; v_TT^*] v_AT^T, which can be rearranged to solve for vAT, as [Eq. 14] ... only a minor part of the complex edge kernel is considered that depends on the known noise-free measurements, which in the proposed scenario are anchor-to-anchor and target-to-target distances and angles."

    The block column in Eq. (13) is, per Eq. (10), the anchor-to-target column of K = v^* v^T. Its middle block is K3 = v_AT^* v_AT^T, a quadratic form in the very vector the update (14) claims to estimate. Substituting the noiseless definitions of K1, K3, K4^T into (14) gives v_AT = v_AT identically; the 'estimation' is an algebraic rearrangement of the target vector, not an inference from independent AA/TT information. Moreover, the conclusion says the minor is built from AA and TT distances/angles, but AA and TT edges do not determine the AT block; forming K3 requires the AT measurements themselves. Hence the claimed prediction reduces to its own input by construction.

full rationale

The central derivation chain is partially circular. Equation (13) defines the selected kernel minor as the anchor-to-target column, whose interior block K3 = v_AT^* v_AT^T depends on the unknown anchor-to-target edge vector. Equation (14) then 'solves' for that same vector from this minor. In the noiseless limit the update reduces to the identity v_AT = v_AT, so the derivation is tautological unless the minor is formed from external AT measurements. The Abstract and Conclusion explicitly claim the minor is based only on noise-free anchor-to-anchor and target-to-target information, which cannot supply K3; that claim makes the method unobservable. However, the paper also reports Monte-Carlo simulations against an MDS baseline and a CRLB, so the numerical implementation may have independent content if K is actually built from noisy AT measurements. The CRLB is borrowed from the authors' companion paper [25] but is used as an external benchmark, not to force the algorithmic result. The score reflects one self-definitional step in the central derivation, offset by the presence of independently falsifiable simulation results.

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

No fitted parameters are introduced; the method has no free constants. It relies on the rank-one kernel property, on exact knowledge of anchor and target-target geometry, and on several unproven or under-specified steps (K3 availability, iteration convergence, C^{-1} inversion).

assumptions (5)
  • standard math Noiseless complex edge kernel K = v^* v^T has rank one
    Used in Eqs. (10)-(11); follows from v = Cx and x being a single complex coordinate vector, but requires exact complex edge vectors.
  • domain assumption Anchor-to-anchor and target-to-target edge vectors v_AA and v_TT are known without error
    Used in Eqs. (14)-(15); justified by known anchor positions and known rigid body shape, but not by the measurement noise model.
  • ad hoc to paper The minor [K1; K3; K4^T] can be constructed from available measurements
    K3 = v_AT^* v_AT^T depends on the unknown edges v_AT; whether noisy target-to-anchor measurements are allowed to build K3 is never stated (Eq. 13).
  • ad hoc to paper Iteration Eq. (14) converges to the correct v_AT
    No proof is given; the paper refers to [21], but the rigid body variant with additional known v_TT and the K3 block is not covered directly.
  • ad hoc to paper Equation (12) can recover x from v via C^{-1}
    C is a P x T incidence-like matrix with P = T(T-1)/2, so C^{-1} is not defined; a pseudoinverse or MDS reconstruction is needed but not specified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SMDS-based Rigid Body Localization." pith.science (2026). https://pith.science/paper/PBB3QP7D

@misc{pith2026250901223,
  author       = {Pith},
  title        = {Pith review of: SMDS-based Rigid Body Localization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PBB3QP7D}},
  note         = {Machine review of arXiv:2509.01223}
}
read the original abstract

We consider a novel rigid body localization (RBL) method, based only on a set of measurements of the distances, as well as the angles between sensors of the vehicle to the anchor landmark points. A key point of the proposed method is to use a variation of the super multidimensional scaling (SMDS) algorithm, where only a minor part of the complex edge kernel is used, based on the available information, which in the case of RBL is anchor-to-anchor and target-to-target information. Simulation results illustrate the good performance of the proposed technique in terms of mean square error (MSE) of the estimates, compared also to the corresponding Cram\'er-Rao Lower Bound (CRLB).

Figures

Figures reproduced from arXiv: 2509.01223 by the authors.

Figure 1
Figure 1. Illustration of an example RBL scenario. The shape of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. root mean square error (RMSE) of the translation [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. MSE of the rotation estimate of the proposed method [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 20 canonical work pages

  1. [25]

    Fundamental limits of rigid body localization,

    N. F ¨uhrling, I. A. M. Sandoval, G. T. F. de Abreu, G. Seco- Granados, D. G. G., and O. Gonsa, “Fundamental limits of rigid body localization,” 2025. [Online]. Available: https://arxiv.org/abs/ 2507.22573

    arXiv 2025

  2. [21]

    Turbo MRC-SMDS: Low-complexity co- operative localization from hybrid information,

    G. Abreu and A. Ghods, “Turbo MRC-SMDS: Low-complexity co- operative localization from hybrid information,” in 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP) , 2018, pp. 76–80

    work page 2018

  3. [1]

    Recent advances in indoor localization: A survey on theoretical approaches and applications,

    A. Yassin, Y . Nasser, M. Awad, A. Al-Dubai, R. Liu, C. Yuen, R. Raulefs, and E. Aboutanios, “Recent advances in indoor localization: A survey on theoretical approaches and applications,” IEEE Communi- cations Surveys & Tutorials , vol. 19, no. 2, pp. 1327–1346, 2017

    work page 2017

  4. [2]

    M.2160-0: Framework and overall objectives of the future development of IMT for 2030 and beyond,

    ITU-R, International Telecommunication Union - Radiocommunication Sector, “M.2160-0: Framework and overall objectives of the future development of IMT for 2030 and beyond,” Nov. 2023

    work page

  5. [3]

    A survey of fingerprint-based outdoor localiza- tion,

    Q. D. V o and P. De, “A survey of fingerprint-based outdoor localiza- tion,” IEEE Communications Surveys & Tutorials , vol. 18, no. 1, pp. 491–506, 2016

    work page 2016

  6. [4]

    Robust received signal strength indicator (RSSI)-based multitarget localization via gaussian process regression,

    N. F ¨uhrling, H. S. Rou, G. T. F. de Abreu, D. G. G., and O. Gonsa, “Robust received signal strength indicator (RSSI)-based multitarget localization via gaussian process regression,” IEEE Journal of Indoor and Seamless Positioning and Navigation , vol. 1, pp. 104–114, 2023

    work page 2023

  7. [5]

    AOA localization for vehicle-tracking systems using a dual-band sensor array,

    M. A. G. Al-Sadoon, R. Asif, Y . I. A. Al-Yasir, R. A. Abd-Alhameed, and P. S. Excell, “AOA localization for vehicle-tracking systems using a dual-band sensor array,” IEEE Transactions on Antennas and Propagation, vol. 68, no. 8, pp. 6330–6345, 2020

    work page 2020

  8. [6]

    Global and asymptotically efficient localization from range measurements,

    G. Zeng, B. Mu, J. Chen, Z. Shi, and J. Wu, “Global and asymptotically efficient localization from range measurements,” IEEE Transactions on Signal Processing, vol. 70, pp. 5041–5057, 2022

    work page 2022

Show all 25 references
  1. [7]

    Super MDS: Source location from distance and angle information,

    G. T. F. de Abreu and G. Destino, “Super MDS: Source location from distance and angle information,” in 2007 IEEE Wireless Communica- tions and Networking Conference , 2007, pp. 4430–4434

    2007

  2. [8]

    On the generation of Tikhonov variates,

    G. T. F. de Abreu, “On the generation of Tikhonov variates,” IEEE Trans. Communications, vol. 56, no. 7, pp. 1157–1168, Jul. 2008

    2008

  3. [9]

    Multidimensional scaling: I. theory and method,

    W. S. Torgerson, “Multidimensional scaling: I. theory and method,” Psychometrika, vol. 17, no. 4, pp. 401–419, Dec. 1952

    1952

  4. [10]

    T. F. Cox and M. A. A. Cox, Multidimensional Scaling , 2nd ed. Chapman & Hall/CRC, 2000

    2000

  5. [11]

    An investigation and solution of angle based rigid body localization,

    Y . Wang, G. Wang, S. Chen, K. C. Ho, and L. Huang, “An investigation and solution of angle based rigid body localization,” IEEE Transactions on Signal Processing , vol. 68, pp. 5457–5472, 2020

    2020

  6. [12]

    Enabling next-generation V2X perception: Wireless rigid body localization and tracking,

    N. F ¨uhrling, H. S. Rou, G. T. F. de Abreu, D. G. G., and O. Gonsa, “Enabling next-generation V2X perception: Wireless rigid body localization and tracking,” 2024. [Online]. Available: https://arxiv.org/abs/2408.00349

    2024 arXiv

  7. [13]

    Tracking position and orientation of a mobile rigid body,

    S. P. Chepuri, A. Simonetto, G. Leus, and A.-J. van der Veen, “Tracking position and orientation of a mobile rigid body,” in 2013 5th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP) , 2013, pp. 37–40

    2013

  8. [14]

    Nonlinear observer for 3D rigid body motion estimation using doppler measure- ments,

    S. Br ´as, M. Izadi, C. Silvestre, A. Sanyal, and P. Oliveira, “Nonlinear observer for 3D rigid body motion estimation using doppler measure- ments,” IEEE Transactions on Automatic Control , vol. 61, no. 11, pp. 3580–3585, 2016

    2016

  9. [15]

    Tightly-coupled visual-inertial localization and 3-D rigid-body target tracking,

    K. Eckenhoff, Y . Yang, P. Geneva, and G. Huang, “Tightly-coupled visual-inertial localization and 3-D rigid-body target tracking,” IEEE Robotics and Automation Letters , vol. 4, no. 2, pp. 1541–1548, 2019

    2019

  10. [16]

    Multi object tracking for predictive collision avoid- ance,

    B. Gebregziabher, “Multi object tracking for predictive collision avoid- ance,” 2023

    2023

  11. [17]

    A survey on trajectory-prediction methods for autonomous driving,

    Y . Huang, J. Du, Z. Yang, Z. Zhou, L. Zhang, and H. Chen, “A survey on trajectory-prediction methods for autonomous driving,” IEEE Transactions on Intelligent V ehicles, vol. 7, no. 3, pp. 652–674, 2022

    2022

  12. [18]

    Robust egoistic rigid body localization,

    N. F ¨uhrling, G. T. F. de Abreu, D. G. G., and O. Gonsa, “Robust egoistic rigid body localization,” 2025. [Online]. Available: https://arxiv.org/abs/2501.10219

    2025

  13. [19]

    Accurate localization of a rigid body using mul- tiple sensors and landmarks,

    S. Chen and K. C. Ho, “Accurate localization of a rigid body using mul- tiple sensors and landmarks,” IEEE Transactions on Signal Processing , vol. 63, no. 24, pp. 6459–6472, 2015

    2015

  14. [20]

    6D rigid body localization and velocity estimation via gaussian belief propagation,

    N. F ¨uhrling, V . Vizitiv, K. R. R. Ranasinghe, H. S. Rou, G. T. F. de Abreu, D. G. G., and O. Gonsa, “6D rigid body localization and velocity estimation via gaussian belief propagation,” 2024. [Online]. Available: https://arxiv.org/abs/2412.12133

    2024

  15. [22]

    Least-squares estimation of transformation parameters between two point patterns,

    S. Umeyama, “Least-squares estimation of transformation parameters between two point patterns,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 13, no. 4, pp. 376–380, 1991

    1991

  16. [23]

    Estimating 3-D rigid body transformations: a comparison of four major algorithms,

    D. W. Eggert, A. Lorusso, and R. B. Fisher, “Estimating 3-D rigid body transformations: a comparison of four major algorithms,” Machine Vision and Applications , vol. 9, no. 5, pp. 272–290, 1997

    1997

  17. [24]

    Papoulis and S

    A. Papoulis and S. U. Pillai, Probability, Random V ariables and Stochastic Processes, 4th ed. New York, NY: Mc-Graw-Hill, 2002

    2002

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.