pith. sign in

arxiv: 2604.17811 · v1 · submitted 2026-04-20 · 📡 eess.SY · cs.SY· math.OC

Kill-Probability-Maximization Guidance: Breaking from the Miss-Distance-Minimization Paradigm

Liraz Mudrik , Yaakov Oshman This is my paper

Pith reviewed 2026-05-10 04:36 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC
keywords missile guidancesingle-shot kill probabilitySSKPBayesian decision theorydifferential gameswarhead lethalityguidance law modification
0
0 comments X

The pith

A new guidance method maximizes the chance of killing the target instead of minimizing the closest approach distance.

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

Classical guidance laws try to get the interceptor as close as possible to the target, which sets a minimum size for the warhead to be effective. This paper changes the objective to directly raising the single-shot kill probability by folding in the warhead's probabilistic lethality curve. The change is achieved by adjusting existing deterministic differential-game guidance laws through Bayesian decision theory while staying consistent with a generalized separation principle. Monte Carlo runs against both standard and evasive targets show higher kill probabilities than either standard or estimation-aware alternatives.

Core claim

Classical guidance laws aim at minimizing the miss distance, thus implicitly determining the minimum warhead lethality radius required against nominal targets. However, nonnominal targets or scenarios might render the designed warhead insufficient, causing a significant degradation in the single-shot kill probability (SSKP). The proposed guidance methodology shifts the interceptor's objective from minimizing the miss distance to directly maximizing the SSKP, while taking into account the warhead's probabilistic lethality model. Complying with the generalized separation theorem, the new paradigm is based on modifying deterministic differential-game-based guidance laws using Bayesian decision

What carries the argument

Modification of deterministic differential-game-based guidance laws using Bayesian decision theory to incorporate the warhead's probabilistic lethality model while obeying the generalized separation theorem.

If this is right

  • Guidance laws can be adapted to different warhead lethality models by simple insertion of the lethality function into the Bayesian update.
  • Performance improves against nonnominal or maneuvering targets without requiring larger warheads.
  • The structure of proven differential-game laws is preserved, allowing reuse of existing design and stability analysis.
  • Simulations show consistent SSKP gains over both classical and estimation-aware guidance across nominal and nonnominal cases.

Where Pith is reading between the lines

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

  • Interceptor designers could trade reduced warhead mass for the same overall kill effectiveness if lethality is optimized in the loop.
  • The same Bayesian adjustment might apply to other pursuit problems where success probability replaces distance as the figure of merit.
  • Sensor noise or target-state estimation error could be folded directly into the same decision-theoretic update without breaking the separation structure.

Load-bearing premise

The generalized separation theorem still permits optimal modification of deterministic guidance laws when the objective is switched to probabilistic kill maximization.

What would settle it

Monte Carlo trials or a closed-form counterexample in which the Bayesian-modified laws produce lower or equal SSKP than standard miss-distance minimization under the same warhead lethality model.

Figures

Figures reproduced from arXiv: 2604.17811 by Liraz Mudrik, Yaakov Oshman.

Figure 1
Figure 1. Figure 1: Planar engagement geometry. The following assumptions are adopted: both players are modeled as point masses; the interceptor’s path angle and lateral acceleration are known via its own navigation system; both speeds VM and VT are known and time￾invariant; both players possess first-order dynamics with known time constants τM and τT ; the lateral acceleration bounds a max M and a max T are known constants; … view at source ↗
Figure 2
Figure 2. Figure 2: compares the new PLM (18) with a CC model having a lethality radius of RSK = 10 m. The PLM’s parameters, µw = 12.5 m and σw = 2.5 m, are chosen so that the effective lethal radius Reff(1) ≜ µw −σw coincides with RSK, enabling a direct comparison of the two models at the same design radius. The dash￾dotted vertical line marks this common radius, where the PLM attains a kill probability of Pd[Reff(1)] ≈ 0.84… view at source ↗
Figure 3
Figure 3. Figure 3: Game space decomposition (µ > 1). Optimal trajectories, and boundaries of the singular region (solid lines). by D − 1 . The optimal feedback strategies in the regular region are u¯ ∗ (τ ) = ¯v ∗ (τ ) = sgn ¯z(τ ). (29) In the singular region, D0, the optimal strategies of both players are arbitrary. For practical reasons we use a linear, chattering prevention strategy for the pursuer inside the singular re… view at source ↗
Figure 4
Figure 4. Figure 4: Miss distance cost function over a dimensional [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Miss probability cost function over a dimen [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Posterior PDF represented by 1000 equally [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
read the original abstract

Classical guidance laws aim at minimizing the miss distance, thus implicitly determining the minimum warhead lethality radius required against nominal targets. However, nonnominal targets or scenarios might render the designed warhead insufficient, causing a significant degradation in the single-shot kill probability (SSKP). We propose a guidance methodology that shifts the interceptor's objective from minimizing the miss distance to directly maximizing the SSKP, while taking into account the warhead's probabilistic lethality model. Complying with the generalized separation theorem, the new paradigm is based on modifying deterministic differential-game-based guidance laws using Bayesian decision theory. Extensive Monte Carlo simulations demonstrate consistent SSKP improvement over the standard and recently introduced estimation-aware guidance laws, when tested against nominal and nonnominal evasively maneuvering targets.

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 paper proposes shifting interceptor guidance from miss-distance minimization to direct maximization of single-shot kill probability (SSKP) by incorporating a probabilistic warhead lethality model. It modifies deterministic differential-game-based guidance laws via Bayesian decision theory while claiming compliance with the generalized separation theorem. Monte Carlo simulations are presented showing consistent SSKP gains over standard and estimation-aware laws against both nominal and nonnominal evasively maneuvering targets.

Significance. If the separation property is preserved under the SSKP objective and the simulation evidence is reproducible, the work offers a paradigm shift in guidance design with potential to improve effectiveness against uncertain or nonnominal threats. The explicit use of Bayesian decision theory to adapt existing differential-game laws, combined with simulation-based validation, provides a concrete path for incorporating lethality models without discarding classical structures.

major comments (2)
  1. [Methodology] Methodology section: The central claim that the generalized separation theorem continues to hold after replacing the deterministic miss-distance terminal cost with the expected SSKP (integral over the lethality probability density) is asserted but not derived. The probabilistic mapping from miss distance to kill probability may introduce coupling between the state estimate and the control that violates separability; an explicit condition or value-function decomposition confirming separation is required for the modification to be guaranteed optimal.
  2. [Simulation results] Simulation results section: The Monte Carlo evidence for consistent SSKP improvement is load-bearing for the practical claim, yet the lethality model (functional form, parameters, and uncertainty), number of runs, exact modifications applied to the baseline differential-game laws, and any statistical tests for significance are not reported. Without these, the robustness of the reported gains cannot be assessed.
minor comments (2)
  1. [Abstract] The abstract and introduction would benefit from a brief statement of the lethality model functional form (e.g., whether it is a Gaussian or exponential function of miss distance) to allow readers to immediately understand the probabilistic objective.
  2. [Methodology] Notation for the Bayesian update and the modified guidance command should be introduced with a clear table or list of symbols to avoid ambiguity when the deterministic law is altered.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the potential of the SSKP-maximization paradigm. We address each major comment below with clarifications and commit to revisions that strengthen the presentation without altering the core contributions.

read point-by-point responses

  1. Referee: [Methodology] Methodology section: The central claim that the generalized separation theorem continues to hold after replacing the deterministic miss-distance terminal cost with the expected SSKP (integral over the lethality probability density) is asserted but not derived. The probabilistic mapping from miss distance to kill probability may introduce coupling between the state estimate and the control that violates separability; an explicit condition or value-function decomposition confirming separation is required for the modification to be guaranteed optimal.

    Authors: We agree that an explicit derivation is required to substantiate the claim. The manuscript relies on the fact that the expected SSKP is obtained by integrating the lethality model against the posterior distribution of the relative state, which is fully captured by the state estimate; this structure preserves the separation property of the underlying differential game. In the revised manuscript we will add a dedicated subsection providing the value-function decomposition and stating the precise conditions (lethality model depending only on relative state) under which the optimal control remains a function of the estimate alone, thereby addressing the potential coupling concern. revision: yes

  2. Referee: [Simulation results] Simulation results section: The Monte Carlo evidence for consistent SSKP improvement is load-bearing for the practical claim, yet the lethality model (functional form, parameters, and uncertainty), number of runs, exact modifications applied to the baseline differential-game laws, and any statistical tests for significance are not reported. Without these, the robustness of the reported gains cannot be assessed.

    Authors: We concur that full reporting is essential for reproducibility. The revised manuscript will explicitly state the lethality model (functional form and parameters together with any uncertainty representation), the exact number of Monte Carlo runs, the precise modifications applied to the baseline differential-game laws (including the Bayesian update step), and the statistical tests performed (e.g., confidence intervals or significance levels) on the observed SSKP gains. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation builds on external theorem and standard Bayesian modification without self-referential reduction

full rationale

The paper's central step invokes the generalized separation theorem to justify modifying deterministic differential-game guidance laws via Bayesian decision theory for an SSKP objective. This is presented as compliance with an existing result rather than a derivation that reduces to the paper's own fitted parameters or self-cited uniqueness claims. No equations in the abstract or described chain equate the new objective to the input miss-distance laws by construction, nor is the theorem shown to be an author-internal ansatz. Monte Carlo validation is external to the derivation. The approach remains self-contained against the cited theorem and Bayesian framework.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach relies on the generalized separation theorem and standard Bayesian decision theory applied to an existing lethality model; no free parameters, new axioms, or invented entities are detailed in the abstract.

axioms (1)
  • domain assumption Generalized separation theorem
    Invoked to justify modifying deterministic differential-game-based guidance laws with probabilistic elements.

pith-pipeline@v0.9.0 · 5430 in / 1135 out tokens · 51273 ms · 2026-05-10T04:36:41.233752+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

19 extracted references · 19 canonical work pages · 1 internal anchor

  1. [1]

    Zarchan Tactical and Strategic Missile Guidance, Sixth Edition

    P. Zarchan Tactical and Strategic Missile Guidance, Sixth Edition. Wash- ington, DC: American Institute of Aeronautics and Astronau- tics, Inc., 2012

    work page 2012

  2. [2]

    J. Z. Ben-Asher and I. Yaesh Advances in Missile Guidance Theory, ser. Progress in As- tronautics and Aeronautics: V ol. 180. Reston, V A: American Institute of Aeronautics and Astronautics, 1998

    work page 1998

  3. [3]

    I. G. Shaviv and Y . Oshman Estimation-Guided Guidance and Its Implementation via Se- quential Monte Carlo Computation Journal of Guidance, Control, and Dynamics, vol. 40, no. 2, pp. 402–417, 2017

    work page 2017

  4. [4]

    Gutman On Optimal Guidance for Homing Missiles Journal of Guidance, Control, and Dynamics, vol

    S. Gutman On Optimal Guidance for Homing Missiles Journal of Guidance, Control, and Dynamics, vol. 2, no. 4, pp. 296–300, Jul. 1979

    work page 1979

  5. [5]

    Shinar Solution Techniques for Realistic Pursuit-Evasion Games InControl and Dynamic Systems, ser

    J. Shinar Solution Techniques for Realistic Pursuit-Evasion Games InControl and Dynamic Systems, ser. Advances in Theory and Applications, C. T. Leondes, Ed. New York, NY: Academic Press, Jan. 1981, vol. 17, pp. 63–124

    work page 1981

  6. [6]

    Shinar and V

    J. Shinar and V . Turetsky What Happens When Certainty Equivalence is Not Valid?: Is There an Optimal Estimator for Terminal Guidance?IFAC Proceedings Volumes, vol. 36, no. 8, pp. 175–184, Jun. 2003

    work page 2003

  7. [7]

    Shaferman and Y

    V . Shaferman and Y . Oshman Stochastic Cooperative Interception Using Information Sharing Based on Engagement Staggering Journal of Guidance, Control, and Dynamics, vol. 39, no. 9, pp. 2127–2141, Jul. 2016

    work page 2016

  8. [8]

    N. K. Jaiswal Search, Detection and Damage Assessment InMilitary Operations Research: Quantitative Decision Mak- ing, ser. International Series in Operations Research & Man- agement Science, N. K. Jaiswal, Ed. Boston, MA: Springer US, 1997, pp. 13–58

    work page 1997

  9. [9]

    Shinar, V

    J. Shinar, V . Turetsky, and Y . Oshman Integrated Estimation/Guidance Design Approach for Improved Homing Against Randomly Maneuvering Targets Journal of Guidance, Control, and Dynamics, vol. 30, no. 1, pp. 154–161, Jan. 2007

    work page 2007

  10. [10]

    Weiss and T

    M. Weiss and T. Shima Minimum Effort Pursuit/Evasion Guidance with Specified Miss Distance Journal of Guidance, Control, and Dynamics, vol. 39, no. 5, pp. 1069–1079, 2016

    work page 2016

  11. [11]

    Shinar, I

    J. Shinar, I. Forte, and B. Kantor Mixed strategy guidance: A new high-performance missile guidance law Journal of Guidance, Control, and Dynamics, vol. 17, no. 1, pp. 129–135, 1994

    work page 1994

  12. [12]

    Mudrik and Y

    L. Mudrik and Y . Oshman A unified estimation–guidance framework based on Bayesian decision theory Journal of Guidance, Control, and Dynamics, April 2026, Published Online. doi.org/10.2514/1.G009628

    work page doi:10.2514/1.g009628 2026

  13. [13]

    Mudrik and Y

    L. Mudrik and Y . Oshman A comprehensive approach to directly addressing estimation delays in stochastic guidance Journal of Guidance, Control, and Dynamics, 2026, submitted for publication. arXiv:2603.05363

    work page internal anchor Pith review arXiv 2026

  14. [14]

    Striebel Sufficient statistics in the optimum control of stochastic systems Journal of Mathematical Analysis and Applications, vol

    C. Striebel Sufficient statistics in the optimum control of stochastic systems Journal of Mathematical Analysis and Applications, vol. 12, no. 3, pp. 576–592, Dec. 1965

    work page 1965

  15. [15]

    H. S. Witsenhausen Separation of Estimation and Control for Discrete Time Systems Proceedings of the IEEE, vol. 59, no. 11, pp. 1557–1566, Nov. 1971

    work page 1971

  16. [16]

    H. A. P. Blom and E. A. Bloem Exact Bayesian and Particle Filtering of Stochastic Hybrid Systems IEEE Transactions on Aerospace and Electronic Systems, vol. 43, no. 1, pp. 55–70, Jan. 2007

    work page 2007

  17. [17]

    V . Y . Glizer and V . Turetsky Complete Solution of a Differential Game with Linear Dynam- ics and Bounded Controls Applied Mathematics Research Express, vol. 2008, p. abm012, Jan. 2008

    work page 2008

  18. [18]

    P. L. Goethals, G. L. Boylan, and B. R. Cho Broadening the Damage Function in Modeling an Array of Military Applications Military Operations Research, vol. 20, no. 1, pp. 39–63, 2015

    work page 2015

  19. [19]

    H. L. Van Trees Detection, Estimation, and Modulation Theory, Part I: Detec- tion, Estimation, and Linear Modulation Theory. Hoboken, NJ: Wiley, Apr. 2004. 10 IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. XX, No. XX XXXXX 2026

    work page 2004