pith. sign in

arxiv: 2406.19599 · v1 · pith:7OC6CCDMnew · submitted 2024-06-28 · ⚛️ physics.data-an · math.DS

An insightful approach to bearings-only tracking in log-polar coordinates

Athena Helena Xiourouppa , Dmitry Mikhin , Melissa Humphries , John Maclean This is my paper

Pith reviewed 2026-05-24 00:02 UTC · model grok-4.3

classification ⚛️ physics.data-an math.DS
keywords bearings-only trackinglog-polar coordinatesclosed-form expressionsunscented Kalman filternon-Gaussianityownship manoeuvretarget state momentsmean range error
0
0 comments X

The pith

Closed-form expressions for the target state prior in log-polar coordinates after an ownship manoeuvre enable direct substitution into UKF time updates.

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

The paper derives explicit formulas for the mean, covariance, and higher central moments of the target state distribution in log-polar coordinates once the observing platform turns. These formulas are inserted directly into the prediction step of an unscented Kalman filter in place of sigma-point propagation, producing a modified filter whose output statistics match those of the unmodified version in simulations. The third and fourth moments quantify the non-Gaussianity that appears at the turn, and their values can be monitored to tune the initial range-error assumption so that the final mean estimated range error meets a chosen target. This matters for bearings-only problems because the coordinate system choice determines whether the prediction step remains tractable after a manoeuvre.

Core claim

Closed-form expressions are derived for the mean, covariance, and higher-order moments of the target state prior distribution in log-polar coordinates following an ownship manoeuvre. These allow direct substitution into the time update equations of a UKF without sigma-point propagation, producing a CFE-UKF that performs similarly to the original while the third and fourth moments indicate non-Gaussianity. Monitoring these and initialising relative range error appropriately achieves a desired mean estimated range error.

What carries the argument

Closed-form expressions for the mean, covariance, and third and fourth central moments of the target state in log-polar coordinates after an ownship turn; these expressions enable direct substitution into the filter time-update equations.

If this is right

  • The CFE-UKF achieves output statistics similar to the pure UKF when the closed-form expressions replace sigma-point propagation at the ownship turn.
  • The third and fourth central moments obtained from the closed-form expressions quantify the non-Gaussianity of the target state after the manoeuvre.
  • Monitoring the higher moments together with suitable initialisation of relative range error produces a desired mean estimated range error at the filter output.
  • Availability of the higher-order moments opens extensions of the tracker that cannot be performed with a standard UKF.

Where Pith is reading between the lines

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

  • The same derivation technique could be applied to modified polar coordinates to compare the size of the non-Gaussian terms across the two systems.
  • The explicit moments supply a diagnostic that could trigger a switch to a particle filter when non-Gaussianity exceeds a chosen threshold.
  • Because the expressions are analytic, they could support a stability analysis of the filter that does not rely on Monte-Carlo trials.

Load-bearing premise

Direct substitution of the closed-form mean and covariance into the time-update equations accurately represents the state transition and produces performance equivalent to full sigma-point propagation.

What would settle it

Run the CFE-UKF and standard UKF on identical bearings-only trajectories containing an ownship turn; if the mean estimated range error of the CFE-UKF deviates by more than a few percent from the standard UKF after the turn, the substitution does not preserve accuracy.

Figures

Figures reproduced from arXiv: 2406.19599 by Athena Helena Xiourouppa, Dmitry Mikhin, John Maclean, Melissa Humphries.

Figure 1
Figure 1. Figure 1: Visualisation by samples of the post-manoeuvre distri [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Visualisation by samples of the post-manoeuvre distri [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: A diagram of the predicted ownship position, [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Time series of the two non-Gaussianity metrics. The [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Effect of the initial range variance on MRE (a, [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
read the original abstract

The choice of coordinate system in a bearings-only (BO) tracking problem influences the methods used to observe and predict the state of a moving target. Modified Polar Coordinates (MPC) and Log-Polar Coordinates (LPC) have some advantages over Cartesian coordinates. In this paper, we derive closed-form expressions for the target state prior distribution after ownship manoeuvre: the mean, covariance, and higher-order moments in LPC. We explore the use of these closed-form expressions in simulation by modifying an existing BO tracker that uses the UKF. Rather than propagating sigma points, we directly substitute current values of the mean and covariance into the time update equations at the ownship turn. This modified UKF, the CFE-UKF, performs similarly to the pure UKF, verifying the closed-form expressions. The closed-form third and fourth central moments indicate non-Gaussianity of the target state when the ownship turns. By monitoring these metrics and appropriately initialising relative range error, we can achieve a desired output mean estimated range error (MRE). The availability of these higher-order moments facilitates other extensions of the tracker not possible with a standard UKF.

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 / 1 minor

Summary. The paper derives closed-form expressions for the mean, covariance, and higher-order central moments of the target state in log-polar coordinates (LPC) after an ownship maneuver in bearings-only tracking. It proposes the CFE-UKF variant, which substitutes the analytic mean and covariance directly into the existing UKF time-update equations at the maneuver (instead of sigma-point propagation), and reports that this performs similarly to the standard UKF in simulations. The nonzero third- and fourth-order moments are presented as evidence of non-Gaussianity, and the authors suggest monitoring them (along with appropriate initialization of relative range error) to achieve a desired mean estimated range error (MRE).

Significance. If the closed-form expressions are correct and the direct-substitution approach is valid, the work would supply an efficient, analytic handle on maneuver-induced transitions in LPC without full sigma-point propagation, plus explicit higher-moment diagnostics for non-Gaussian effects that are unavailable in a standard UKF. The derivation of the third- and fourth-order moments is a concrete strength, as it makes the non-Gaussianity claim falsifiable and opens the door to moment-based extensions.

major comments (2)
  1. [Abstract] Abstract: the claim that the CFE-UKF 'performs similarly to the pure UKF' rests on unspecified Monte-Carlo runs with no reported quantitative metrics (RMSE, MRE values, number of trials, scenario parameters, or error bars), so the central assertion that direct substitution of the closed-form mean/covariance is adequate cannot be evaluated from the given evidence.
  2. [Abstract] Abstract (time-update modification): the paper explicitly derives nonzero third- and fourth-order central moments after the turn, confirming that the prior is non-Gaussian, yet the CFE-UKF replaces sigma-point propagation with direct insertion of only the first two moments; no analytic argument is supplied that the UKF update equations remain accurate when the distribution deviates from Gaussianity, which is the load-bearing assumption for the method.
minor comments (1)
  1. The abstract refers to 'verifying the closed-form expressions' via simulation similarity, but without any tabulated results or figure references this verification step is difficult to assess.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the abstract and the underlying assumptions of the CFE-UKF. We address each major comment below and outline the revisions that will be made.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that the CFE-UKF 'performs similarly to the pure UKF' rests on unspecified Monte-Carlo runs with no reported quantitative metrics (RMSE, MRE values, number of trials, scenario parameters, or error bars), so the central assertion that direct substitution of the closed-form mean/covariance is adequate cannot be evaluated from the given evidence.

    Authors: The full manuscript reports Monte Carlo results (including RMSE, MRE, trial counts, and scenario parameters) that support the performance claim. We agree the abstract is insufficiently quantitative on this point. We will revise the abstract to include the key metrics, number of trials, and scenario details so the central assertion can be evaluated directly from the abstract. revision: yes

  2. Referee: [Abstract] Abstract (time-update modification): the paper explicitly derives nonzero third- and fourth-order central moments after the turn, confirming that the prior is non-Gaussian, yet the CFE-UKF replaces sigma-point propagation with direct insertion of only the first two moments; no analytic argument is supplied that the UKF update equations remain accurate when the distribution deviates from Gaussianity, which is the load-bearing assumption for the method.

    Authors: The nonzero higher-order moments correctly establish that the prior is non-Gaussian. The CFE-UKF substitutes the exact analytic mean and covariance (rather than sigma-point estimates) into the existing UKF time-update equations; this is still an approximation because the UKF measurement update treats the prior as Gaussian. We will add a dedicated paragraph in the revised manuscript providing the rationale: the closed-form expressions supply the exact first two moments of the (non-Gaussian) prior, which is at least as accurate as the moment estimates obtained by sigma-point propagation through the maneuver, while the subsequent UKF steps remain unchanged. We will also note the empirical similarity observed in the simulations as supporting evidence and explicitly state the limitation that higher-order moments are not propagated. revision: partial

Circularity Check

0 steps flagged

No circularity: closed-form moments derived from first principles and verified externally

full rationale

The paper derives analytic expressions for the mean, covariance, and higher-order central moments of the target state in log-polar coordinates directly from the coordinate transformation and ownship manoeuvre geometry. These expressions are then substituted into an existing UKF time-update step and compared against full sigma-point propagation in Monte-Carlo simulations. No step reduces a claimed prediction to a fitted parameter by construction, no load-bearing uniqueness theorem is imported via self-citation, and the non-Gaussianity result is presented as an independent observation rather than used to justify the Gaussian approximation. The derivation chain is therefore self-contained against external simulation benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available, preventing identification of specific free parameters, axioms, or invented entities; none are explicitly mentioned in the provided text.

pith-pipeline@v0.9.0 · 5738 in / 1170 out tokens · 30867 ms · 2026-05-24T00:02:41.090156+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

14 extracted references · 14 canonical work pages

  1. [1]

    Aidala and S

    V . Aidala and S. Hammel. Utilization of modified polar coordinates for bearings-only tracking. IEEE Transactions on Automatic Control , 28(3):283–294, March 1983

    work page 1983

  2. [2]

    Hoelzer, G

    Hans D. Hoelzer, G. W. Johnson, and A. O. Cohen. Modified Polar Coordinates - The Key to Well Behaved Bearings Only Ranging. Technical Report IBM IR&D Report 78-M19-0001A, IBM Federal Systems Division, 1978

    work page 1978

  3. [3]

    Algorithm of modified polar coordinates UKF for bearings-only target tracking

    Dan Wang, Hongyan Hua, and Haiwang Cao. Algorithm of modified polar coordinates UKF for bearings-only target tracking. In 2010 2nd International Conference on Future Computer and Communication , volume 3, pages V3–557–V3–560, 2010

    work page 2010

  4. [4]

    Brehard and J

    T. Brehard and J. R. Le Cadre. Closed-form posterior Cram ´er-Rao bounds for bearings-only tracking. IEEE Transactions on Aerospace and Electronic Systems , 42(4):1198–1223, October 2006

    work page 2006

  5. [5]

    Angle-only filtering in 3D using Modified Spherical and Log Spherical Coordinates

    Mahendra Mallick, Sanjeev Arulampalam, Lyudmila Mihaylova, and Yanjun Yan. Angle-only filtering in 3D using Modified Spherical and Log Spherical Coordinates. In Proceedings of the 14th International Conference on Information Fusion (FUSION) , pages 1–8, Chicago, IL, July 2011. IEEE

    work page 2011

  6. [6]

    Kalman Filtering and Neural Networks

    Simon Haykin, editor. Kalman Filtering and Neural Networks . John Wiley & Sons, Inc., October 2001

    work page 2001

  7. [7]

    E. A. Wan and R. van der Merwe. The unscented Kalman filter for nonlinear estimation. In Proceedings of the IEEE 2000 Adaptive Systems for Signal Processing, Communications, and Control Symposium (Cat. No.00EX373). IEEE, 2000

    work page 2000

  8. [8]

    P. L. Houtekamer and Herschel L. Mitchell. Data assimilation using an ensemble kalman filter technique. Monthly Weather Review, 126(3):796– 811, March 1998

    work page 1998

  9. [9]

    Stochastics III lecture notes

    V olker Schmidt. Stochastics III lecture notes. Ulm University, 2012. 11 E x3 1 = − 3 4 ∆v0 3 cos (µ0 − 3 σ30) e(−3 µ3− 1 2 σ00+ 9 2 σ33) − 3 4 ∆v0∆v1 2 cos (µ0 − 3 σ30) e(−3 µ3− 1 2 σ00+ 9 2 σ33) − 1 4 ∆v0 3 cos (3µ0 − 9 σ30) e(−3 µ3− 9 2 σ00+ 9 2 σ33) + 3 4 ∆v0∆v1 2 cos (3µ0 − 9 σ30) e(−3 µ3− 9 2 σ00+ 9 2 σ33) + 3 4 ∆v0 2∆v1e(−3 µ3− 9 2 σ00+ 9 2 σ33) si...

    work page 2012

  10. [10]

    The Normal Distribution: Characterizations with Applications

    Wlodzimierz Bryc. The Normal Distribution: Characterizations with Applications. Springer-Verlag, 1995

    work page 1995

  11. [11]

    SageMath, the Sage Mathematics Software System (Version 9.8) , 2023

    The Sage Developers, William Stein, David Joyner, David Kohel, John Cremona, and Burc ¸in Er¨ocal. SageMath, the Sage Mathematics Software System (Version 9.8) , 2023. https://www.sagemath.org

    work page 2023

  12. [12]

    Kalman and Bayesian Filters in Python

    Roger Labbe. Kalman and Bayesian Filters in Python. https://github. com/rlabbe/Kalman-and-Bayesian-Filters-in-Python, 2014

    work page 2014

  13. [13]

    Rong Li, and Thiagalingam Kirubarajan

    Yaakov Bar-Shalom, X. Rong Li, and Thiagalingam Kirubarajan. Esti- mation with Applications to Tracking and Navigation . John Wiley and Sons, 2001

    work page 2001

  14. [14]

    A novel Gaussian Mixture approxi- mation for nonlinear estimation

    Renato Zanetti and Kirsten Tuggle. A novel Gaussian Mixture approxi- mation for nonlinear estimation. In 2018 21st International Conference on Information Fusion (FUSION) . IEEE, July 2018. BIOGRAPHY SECTION Athena Helena Xiourouppa Having completed a Bachelor of Mathematical Sciences (Advanced) at the University of Adelaide, Athena is pursuing postgradua...

    work page 2018