Pith. sign in

REVIEW 4 major objections 4 minor 29 references

Privacy-Preserving Fusion for Multi-Sensor Systems Under Multiple Packet Dropouts

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a simple encoding step makes an eavesdropper's state estimate diverge while the legitimate fusion filter stays accurate, even when sensor packets are dropped at different rates.

desk verdict A credible multi-sensor extension of an encoding-based privacy scheme, but the confidentiality theorem is vacuous as stated and the mean-divergence proof fails for zero-mean initial states. read the letter →

arxiv 2507.13286 v2 pith:OBK2U23U submitted 2025-07-17 eess.SY cs.SY

Jie Huang , Jason J. R. Liu , Xiao He This is my paper
classification eess.SYcs.SY
keywords privacy-preservingfusionestimationsecurestateencoding-basedmechanismmultiplepacketdropoutsmodifiedalgebraicRiccatiequationeavesdroppingperfectsecrecywirelesssensornetworks
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

Wireless sensor networks broadcast measurements where an eavesdropper can listen, and each sensor's packets may be dropped independently. This paper claims that both problems can be handled together by encoding each measurement before transmission: the legitimate fusion center can still decode and filter, while the eavesdropper's mean state estimate is driven to diverge. The key is a per-sensor encoding parameter $a_i > 1$ that makes the eavesdropper's decoding error grow exponentially once it misses a single reference packet, combined with a centralized filter whose error covariance is bounded by a modified algebraic Riccati equation when total channel capacity exceeds the system's topological entropy. The practical payoff would be confidentiality without encryption and without sacrificing estimation accuracy on the legitimate side.

What carries the argument

The load-bearing object is the encoding-decoding pair defined by equations (6)-(8): each sensor shifts its measurement by $(a_i)^{k - t_{i,k}}$ times the last successfully transmitted measurement, quantizes the difference, and scales it by $1/s$; the legitimate decoder adds the same reference term back. Because the eavesdropper missed the reference packet at the critical event, its decoded value is polluted by a term that grows like $(a_i)^{k - \bar{k}_i - 1}$. The legitimate side is analyzed through the modified algebraic Riccati equation $g_{\{\bar{\gamma}_i,\delta_{N_i}\}}(X)$, whose Hadamard-product form encodes the different dropout probabilities of the individual channels; convergence of this recursion gives the boundedness condition.

What would settle it

In the paper's three-tank simulation with $a_i > 1$, set the channel process so that $\gamma_{i,\bar{k}}=1$ and $\gamma^e_{i,\bar{k}}=0$ with probability one at some $\bar{k}$, and $\gamma^e_{i,k}=1$ for every later $k$; if the eavesdropper's Monte-Carlo mean estimation error remains bounded rather than diverging to infinity, Theorem 4's central claim is false.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 4: if at least one encoding parameter satisfies $a_i > 1$ and the event that the authorized channel transmits while the wiretap channel fails occurs with probability one at some time, then the eavesdropper's expected estimation error diverges, $\lim_{k\to\infty} \|\mathbb{E}[\tilde{x}^e_{k|k}]\| = \infty$. The mechanism is the reference term $(a_i)^{k - \bar{k}_i - 1}$ in the eavesdropper's decoding error, which grows without bound after the reference measurement is missed. On the legitimate side, Theorem 2 claims the expected error covariance is bounded above by the sequence generated by a modified algebraic Riccati equation, and Corollary 1 states that this bound converges whenever the total channel capacity exceeds the topological entropy of the system matrix $A$. Together these two results are meant to establish the paper's definition of secrecy: bounded error for the legitimate user, divergent error for the eavesdropper.

Load-bearing premise

The secrecy conclusion rests on the critical event—one authorized transmission while the wiretap channel fails—occurring with probability one at some time, and the proof tracks that single worst-case path rather than a general stationary random channel; if that event cannot be guaranteed, the divergence result is not established by this paper.

Editorial extensions

If this is right

  • With at least one encoding parameter $a_i > 1$ and one probability-one critical event, no estimator that fuses the wiretapped data can keep its mean error bounded, so confidentiality is a structural consequence of the encoding, not just a statistical heuristic.
  • The legitimate filter remains a recursive Kalman-type estimator, so the privacy mechanism adds only local encoding and decoding operations rather than expensive cryptographic computation.
  • When the total channel capacity exceeds the topological entropy of $A$, the modified algebraic Riccati equation gives a convergent upper bound on the legitimate user's error covariance, yielding a quantitative design target for packet arrival probabilities.
  • The simulations show that the eavesdropper's mean-square error diverges for any tested group with at least one $a_i > 1$, while the legitimate user's MSE is barely affected by $a_i$ and improves as the quantization step $\delta_i$ shrinks.

Reading between the lines

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

  • Beyond the paper's probability-one worst-case event, a stochastic version with the critical event occurring at random times with probability less than one is left open; if the divergence rate depends on how often such events recur, a weaker but still useful secrecy guarantee could be formulated.
  • A smarter eavesdropper could try to reconstruct the missing reference from other sensors' channels or from the quantized-output statistics; the paper assumes the eavesdropper fuses all received data but does not actively exploit cross-channel correlation to cancel the reference term.
  • The capacity condition in Corollary 1 is stated for the total capacity across channels and is proven by decomposing the system into observable modes; a natural design extension the paper does not spell out is to allocate more channel capacity to sensors that observe modes with larger entropy.
  • The same encoding mechanism could be transferred to event-triggered or asynchronous fusion by replacing the fixed reference time $t_{i,k}$ with a dropout-dependent running estimate, but the error growth analysis would need to be reworked in that setting.
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 / 4 minor

Summary. The paper proposes a privacy-preserving fusion estimation (PPFE) scheme for multi-sensor systems in which measurements are encoded before transmission over channels subject to Bernoulli packet dropouts and eavesdropping. It develops a centralized fusion filter, derives a modified algebraic Riccati equation (MARE) intended to bound the legitimate user's error covariance, and claims that the eavesdropper's mean estimation error diverges when at least one encoding parameter satisfies a_i > 1. A three-tank simulation is used to illustrate the trade-off between the legitimate user's mean-square error and the eavesdropper's error.

Significance. If the claims were established, the paper would provide a useful extension of encoding-based privacy mechanisms to multi-sensor fusion with heterogeneous packet-dropout probabilities. The problem is timely, the separation of legitimate-user boundedness from eavesdropper divergence is a sensible framing, and the capacity condition in Corollary 1 is a concrete design criterion. However, the central theoretical guarantees are not rigorously established: the secrecy theorem is based on an event whose probability is zero or strictly less than one under the paper's own channel model, and the mean-divergence proof fails for zero-mean initial states. The legitimate-user boundedness proof also rests on unproved structural assumptions about the MARE and on an unjustified distortion-rate bound. The paper does not provide machine-checked proofs or reproducible code; its value as it stands is mainly in the problem formulation and the numerical illustration, not in the stated guarantees.

major comments (4)
  1. [Section III-C, Theorem 4 (Eq. (60))] The hypothesis Pr{γ_{i,bar{k}_i}=1, γ^e_{i,bar{k}_i}=0}=1 is incompatible with the channel model in (2)-(3). Since γ^e_{i,bar{k}_i} is Bernoulli with Pr{γ^e_{i,bar{k}_i}=1}=bar{γ}^e_i ∈ (0,1], we have Pr{γ^e_{i,bar{k}_i}=0}=1-bar{γ}^e_i, which is strictly less than 1 whenever bar{γ}^e_i ∈ (0,1) and is 0 when bar{γ}^e_i=1. Thus the event in Theorem 4 has probability zero or strictly less than one for every admissible parameter choice. Similarly, Lemma 3 conditions on E_{W_i}={γ^e_{i,k}=1 for all k>bar{k}_i}, a tail event with probability zero for a Bernoulli wiretap channel with bar{γ}^e_i<1. Consequently, the claimed divergence result covers no sample path of the stated stochastic model, and the confidentiality condition in Definition 1(ii) is not established.
  2. [Section III-C, Lemma 3 (Eq. (61))] The proof asserts without argument that E[bar{e}_{i,bar{k}_i+1}-e_{i,bar{k}_i+1}s] does not converge to zero. If the initial state has zero mean, then E[y_{i,k}]=0 for all k; because the quantization error has zero mean and the encoding-decoding recursion is affine in expectation, E[bar{y}_{i,k}]=0 for both the legitimate receiver and the eavesdropper. The reference-mismatch term in (61) is then zero, so (61) gives E[bar{e}_{i,k}]=0 for all k and the mean recursion (57) stays at zero. The paper neither excludes zero-mean initial states nor provides a condition under which the reference term is nonzero. Therefore the claimed divergence lim_{k→∞} ||E[tilde{x}^e_{k|k}]|| = ∞ does not follow, and this is a load-bearing failure for the paper's secrecy guarantee.
  3. [Section III-B, Lemma 1 (Eq. (22))] The key bound E[e_{i,k} e_{i,k}^T] ≤ δ_{N_i}(C_i Σ_{k|k-1} C_i^T + R_i), with δ_{N_i}→0 as N_i→∞, is asserted without proof. The PPM (6)-(8) fixes a step size δ_i and the uniform-quantization rule gives Var(e_{i,k}(ℓ)) ≤ δ_i^2/4; no N_i appears in the encoding mechanism, and no argument shows that the quantization-error covariance is dominated by the innovation covariance with an arbitrarily small factor. Since Lemma 1 supplies the matrix V in (20)-(21) and the upper bound bar{U}_k in (27), the boundedness result for the legitimate user in Theorem 2 rests on this unproved distortion-rate assumption.
  4. [Section III-B, Theorem 2 (Eqs. (29), (35)-(38))] The induction step uses the assertions that g_{...}(X) is concave and monotonically increasing in X and then applies Jensen's inequality as E[g(U^-_k)] ≤ g(E[U^-_k]). No proof or citation is given for these properties of the MARE with the Hadamard product in (29). In addition, W_k and H_k in (26) are defined using the running covariance Σ_{k|k-1}, so it is not clear that the operator in (29) is a well-defined function of the argument X alone as required for the deterministic comparison sequence V_k. Thus the boundedness proof for the legitimate user is not rigorous as presented.
minor comments (4)
  1. [Eq. (26)] The notation 'W_kW_k ≜ ...' is dimensionally inconsistent: the right-hand side is a scalar multiple of an identity matrix, so the left-hand side should be W_k^2 or the definition of W_k itself. This affects the readability of (27).
  2. [Theorem 3 and Eq. (37)] The symbol bar{Γ} is defined inconsistently: the theorem statement uses bar{Γ} ≜ diag{γ_{i,k} I_{d_{y_i}}}, while Eq. (37) defines bar{Γ} ≜ diag{bar{γ}_i I_{d_{y_i}}}. This ambiguity makes condition (39) difficult to interpret.
  3. [Section III-C, Lemma 2] The lower bound (K^e_k)^T K^e_k ≥ ... uses λ_min{Q}, which is zero when Q is only positive semidefinite, and the proof assumes without justification that the eavesdropper's prediction covariance is bounded. The lemma should be restated with explicit sufficient conditions.
  4. [Section IV] Definition 1(ii) concerns the divergence of the mean estimation error, but the simulations plot MSE; reporting the mean estimation error would make the empirical support for the theoretical confidentiality claim more direct.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the PPFE result is derived from the stated encoding and channel model rather than from its own conclusion, despite unresolved mathematical gaps in the secrecy proof.

full rationale

The paper's derivation chain is not circular. The legitimate-user filter (Theorem 1) and the modified algebraic Riccati equation bound (Theorems 2-3) are obtained from the stated process model, channel model, and encoding/decoding definitions using standard Kalman-filter and Riccati arguments; Corollary 1 invokes the external Wonham-decomposition result [25], not a self-citation. The confidentiality analysis (Lemmas 2-3 and Theorem 4) is developed in-paper from the encoding recursion in Eq. (61), with the exponential factor (a_i)^{k-...} generating the claimed divergence, and Lemma 4 from the authors' preprint [28] is used only as an auxiliary conditional-expectation inequality rather than as a substitute for the claimed secrecy result. The cited encoding structure [22] is explicitly identified as the starting point and is then extended to the multi-sensor dropout setting; no fitted parameter is relabeled as a prediction. There are genuine mathematical concerns in the secrecy proof, notably the unproved assertion that E[bar(e)_{i,bar(k)_i+1}-e_{i,bar(k)_i+1}s] is nonzero for zero-mean initial states and the probability-one hypothesis in Theorem 4 being inconsistent with the Bernoulli channel model, but these are correctness risks rather than circular reductions. No step quoted from the paper makes the output equivalent to its input by construction.

Assumptions & free parameters 4 free parameters · 3 assumptions · 2 invented entities

The central claims depend on several chosen encoding parameters, an approximate Gaussian distribution assumption, an invoked distortion bound, and a concavity property of the Riccati operator. The MARE and the critical-event structure are introduced by the paper as tools for stating the results; they have partial independent verification only through the simulation.

free parameters (4)
  • a_i = simulation: a1=0.5, a2=0.5 or 5, a3=5 or 10
    Encoding divergence rate for each channel. Chosen by hand to illustrate the tradeoff; the theory requires a_i > 1. In the simulation groups, some a_i are set to 0.5, violating the theoretical condition for those channels.
  • s = 1 in simulation
    Scaling parameter in the encoding/decoding mechanism. The paper notes a tradeoff in its design but does not provide a systematic selection rule.
  • delta_i = 0.01 or varied as 0.1, 0.01, 0.001 in simulation
    Quantization step size. This finite precision parameter controls the legitimate user's decoding error and is tuned in the simulation.
  • eta_i = not specified
    Positive scalars introduced in Lemma 1 to bound the cross-correlation term; their values are never chosen or discussed.
assumptions (3)
  • domain assumption The decoded measurement distribution is approximately Gaussian so a Kalman-like recursive filter can be applied.
    Section III-A states that the distribution closely approximates a Gaussian as the number of encoding levels increases, citing [23], [24]. This approximation justifies the filter structure.
  • domain assumption The distortion rate \delta_{N_i} exists such that the decoding error covariance is bounded by \delta_{N_i}(C_i \Sigma_{k|k-1} C_i^T + R_i).
    Equation (22) in Lemma 1 assumes an upper bound that depends on the prediction error covariance itself. The existence and time-uniformity of \delta_{N_i} are not rigorously established.
  • ad hoc to paper The Riccati operator g is concave and monotonically increasing in X.
    Theorem 2 uses these properties to apply Jensen's inequality for the induction step. The monotonicity and concavity are asserted without a full proof and are central to the boundedness argument.
invented entities (2)
  • Modified algebraic Riccati equation g(X) with Hadamard product structure
    purpose: Generates an upper bound sequence for the legitimate user's expected error covariance under multi-channel random dropouts and encoding noise.
    The MARE is constructed as a variance recursion and its boundedness is proven by connecting to known results [25], [26], but it is not independently verified by a separate estimator or measurement.
  • Critical event E_{C_i} and worst-case wiretap event E_{W_i} independent evidence
    purpose: A particular dropout pattern needed to prove the eavesdropper error divergence.
    The events are defined in terms of the channel Bernoulli processes, so their probability is falsifiable and testable in simulation; however, the paper's guarantee only covers the event occurring with probability one.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Privacy-Preserving Fusion for Multi-Sensor Systems Under Multiple Packet Dropouts." pith.science (2026). https://pith.science/paper/OBK2U23U

@misc{pith2026250713286,
  author       = {Pith},
  title        = {Pith review of: Privacy-Preserving Fusion for Multi-Sensor Systems Under Multiple Packet Dropouts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OBK2U23U}},
  note         = {Machine review of arXiv:2507.13286}
}
read the original abstract

Wireless sensor networks (WSNs) are critical components in modern cyber-physical systems, enabling efficient data collection and fusion through spatially distributed sensors. However, the inherent risks of eavesdropping and packet dropouts in such networks pose significant challenges to secure state estimation. In this paper, we address the privacy-preserving fusion estimation (PPFE) problem for multi-sensor systems under multiple packet dropouts and eavesdropping attacks. To mitigate these issues, we propose a distributed encoding-based privacy-preserving mechanism (PPM) within a control-theoretic framework, ensuring data privacy during transmission while maintaining the performance of legitimate state estimation. A centralized fusion filter is developed, accounting for the coupling effects of packet dropouts and the encoding-based PPM. Boundedness conditions for the legitimate user's estimation error covariance are derived via a modified algebraic Riccati equation. Additionally, by demonstrating the divergence of the eavesdropper's mean estimation error, the proposed PPFE algorithm's data confidentiality is rigorously analyzed. Simulation results for an Internet-based three-tank system validate the effectiveness of the proposed approach, highlighting its potential to enhance privacy without compromising estimation accuracy.

Figures

Figures reproduced from arXiv: 2507.13286 by the authors.

Figure 1
Figure 1. The fusion estimation architecture in presence of an eavesdropper [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. MSE of the legitimate user’s and eavesdropper’s estimation under [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. MSE of the legitimate user’s estimation under different encoding [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages

  1. [22]

    Privacy-preserving state estimation with unreliable channels,

    J. Huang, C. Gao, and X. He, “Privacy-preserving state estimation with unreliable channels,” ISA Transactions, vol. 127, pp. 4–12, 2022

    work page 2022

  2. [28]

    Recursive Privacy-Preserving Estimation Over Markov Fading Channels

    J. Huang, F. Jia, and X. He, “Recursive privacy-preserving estimation over markov fading channels,” 2025. [Online]. Available: https: //arxiv.org/abs/2506.02725

    work page Pith review arXiv 2025

  3. [1]

    Distributed covariance intersection fusion estimation for cyber-physical systems with commu- nication constraints,

    B. Chen, G. Hu, D. W. C. Ho, and L. Yu, “Distributed covariance intersection fusion estimation for cyber-physical systems with commu- nication constraints,” IEEE Transactions on Automatic Control , vol. 61, no. 12, pp. 4020–4026, 2016

    work page 2016

  4. [2]

    A survey on data fusion approaches in iot-based smart cities: Smart applications, taxonomies, challenges, and future research directions,

    B. Cengiz, I. Y . Adam, M. Ozdem, and R. Das, “A survey on data fusion approaches in iot-based smart cities: Smart applications, taxonomies, challenges, and future research directions,” Information Fusion, vol. 121, p. 103102, 2025. [Online]. Available: https: //www.sciencedirect.com/science/article/pii/S1566253525001757

    work page 2025

  5. [3]

    Distributed fusion estimation for unstable systems with quantized innovations,

    B. Xiang, B. Chen, and L. Yu, “Distributed fusion estimation for unstable systems with quantized innovations,” IEEE Transactions on Systems, Man, and Cybernetics: Systems , vol. 51, no. 10, pp. 6381–6387, 2021

    work page 2021

  6. [4]

    Distributed robust fusion estimation with application to state monitoring systems,

    B. Chen, G. Hu, D. W. C. Ho, W.-A. Zhang, and L. Yu, “Distributed robust fusion estimation with application to state monitoring systems,” IEEE Transactions on Systems, Man, and Cybernetics: Systems , vol. 47, no. 11, pp. 2994–3005, 2017

    work page 2017

  7. [5]

    Reliable data fusion of hierarchical wireless sensor networks with asynchronous measurement for greenhouse monitoring,

    X. Bai, Z. Wang, L. Sheng, and Z. Wang, “Reliable data fusion of hierarchical wireless sensor networks with asynchronous measurement for greenhouse monitoring,” IEEE Transactions on Control Systems Technology, vol. 27, no. 3, pp. 1036–1046, 2019

    work page 2019

  8. [6]

    Towards invariant extended kalman filter-based resilient distributed state estimation for moving robots over mobile sensor networks under deception attacks,

    C. Zhang, J. Qin, C. Yan, Y . Shi, Y . Wang, and M. Li, “Towards invariant extended kalman filter-based resilient distributed state estimation for moving robots over mobile sensor networks under deception attacks,” Automatica, vol. 159, p. 111408, 2024. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0005109823005757

    work page 2024

Show all 29 references
  1. [7]

    Optimal dos attack scheduling for cyber–physical systems with channel hopping scheme,

    X.-H. Liu, J. Dong, and G.-H. Yang, “Optimal dos attack scheduling for cyber–physical systems with channel hopping scheme,” IEEE Transac- tions on Systems, Man, and Cybernetics: Systems , vol. 54, no. 1, pp. 193–200, 2024

    2024

  2. [8]

    Fully distributed secure consensus control for cyber-physical systems against actuator fault and denial-of-service attack,

    Y . Long, X. Yang, T. Li, and H. Liang, “Fully distributed secure consensus control for cyber-physical systems against actuator fault and denial-of-service attack,” IEEE Transactions on Automation Science and Engineering, pp. 1–1, 2025

    2025

  3. [9]

    Cyber-security enhancement of networked control systems using homomorphic encryption,

    K. Kogiso and T. Fujita, “Cyber-security enhancement of networked control systems using homomorphic encryption,” in 2015 54th IEEE Conference on Decision and Control (CDC) , 2015, pp. 6836–6843

    2015

  4. [10]

    Differentially private con- sensus control for discrete-time multiagent systems: Encoding-decoding schemes,

    C. Gao, Z. Wang, X. He, Y . Liu, and D. Yue, “Differentially private con- sensus control for discrete-time multiagent systems: Encoding-decoding schemes,” IEEE Transactions on Automatic Control , vol. 69, no. 8, pp. 5554–5561, 2024

    2024

  5. [11]

    Privacy- preserving platooning control of vehicular cyber–physical systems with saturated inputs,

    D. Pan, D. Ding, X. Ge, Q.-L. Han, and X.-M. Zhang, “Privacy- preserving platooning control of vehicular cyber–physical systems with saturated inputs,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 53, no. 4, pp. 2083–2097, 2023

    2023

  6. [12]

    Privacy preserving for switched systems under robust data-driven predictive control,

    Y . Qi, S. Guo, R. Chi, Y . Tang, and Z. Qu, “Privacy preserving for switched systems under robust data-driven predictive control,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 55, no. 1, pp. 480–490, 2025

    2025

  7. [13]

    Privacy-preserving fusion estimation for power system based on channel gain encryption*,

    Q. Ge, J. Liu, N. Ding, J. Xu, S. Zhan, and D. Xu, “Privacy-preserving fusion estimation for power system based on channel gain encryption*,” in 2024 7th International Symposium on Autonomous Systems (ISAS) , 2024, pp. 1–6

    2024

  8. [14]

    Guaranteeing differential privacy in distributed fusion estimation,

    X. Yan, B. Chen, Y . Zhang, and L. Yu, “Guaranteeing differential privacy in distributed fusion estimation,” IEEE Transactions on Aerospace and Electronic Systems, vol. 59, no. 3, pp. 3416–3423, 2023

    2023

  9. [15]

    Distributed confidentiality fu- sion estimation against eavesdroppers,

    X. Yan, Y . Zhang, D. Xu, and B. Chen, “Distributed confidentiality fu- sion estimation against eavesdroppers,” IEEE Transactions on Aerospace and Electronic Systems , vol. 58, no. 4, pp. 3633–3642, 2022

    2022

  10. [16]

    State estimation with secrecy against eavesdroppers,

    A. Tsiamis, K. Gatsis, and G. J. Pappas, “State estimation with secrecy against eavesdroppers,” IFAC PapersOnLine, vol. 50, no. 1, pp. 8385– 8392, 2017

    2017

  11. [17]

    On remote state estimation in the presence of an eavesdropper,

    A. S. Leong, D. E. Quevedo, D. Dolz, and S. Dey, “On remote state estimation in the presence of an eavesdropper,” IFAC PapersOnLine, vol. 50, no. 1, pp. 7339–7344, 2017

    2017

  12. [18]

    Remote state estimation with privacy against eavesdroppers,

    M. J. Crimson, J. M. Kennedy, and D. E. Quevedo, “Remote state estimation with privacy against eavesdroppers,” IFAC-PapersOnLine, vol. 56, no. 2, pp. 8363–8368, 2023, 22nd IFAC World Congress. [Online]. Available: https://www.sciencedirect.com/science/article/pii/ S2405896323014118

    2023

  13. [19]

    Encryption-decryption-based state estimation with multirate measurements against eavesdroppers: A recursive minimum-variance approach,

    L. Zou, Z. Wang, B. Shen, and H. Dong, “Encryption-decryption-based state estimation with multirate measurements against eavesdroppers: A recursive minimum-variance approach,” IEEE Transactions on Auto- matic Control, vol. 68, no. 12, pp. 8111–8118, 2023

    2023

  14. [20]

    State estimation codes for perfect secrecy,

    A. Tsiamis, K. Gatsis, and G. J. Pappas, “State estimation codes for perfect secrecy,” in 2017 IEEE 56th Annual Conference on Decision and Control (CDC) , 2017, pp. 176–181

    2017

  15. [21]

    State-secrecy codes for stable systems,

    ——, “State-secrecy codes for stable systems,” in 2018 Annual American Control Conference (ACC), 2018, pp. 171–177

    2018

  16. [23]

    Soi-kf: Distributed kalman filtering with low-cost communications using the sign of inno- vations,

    A. Ribeiro, G. B. Giannakis, and S. I. Roumeliotis, “Soi-kf: Distributed kalman filtering with low-cost communications using the sign of inno- vations,” IEEE Transactions on Signal Processing , vol. 54, no. 12, pp. 4782–4795, 2006

    2006

  17. [24]

    Recursive quantized state estimation of discrete-time linear stochastic systems,

    K. You, Y . Zhao, and L. Xie, “Recursive quantized state estimation of discrete-time linear stochastic systems,” in 2009 7th Asian Control Conference, 2009, pp. 170–175

    2009

  18. [25]

    Multi-sensor kalman filtering with intermittent measurements,

    C. Yang, J. Zheng, X. Ren, W. Yang, H. Shi, and L. Shi, “Multi-sensor kalman filtering with intermittent measurements,” IEEE Transactions on Automatic Control, vol. 63, no. 3, pp. 797–804, 2018

    2018

  19. [26]

    On the existence of a stabilizing solution of modified algebraic Riccati equations in terms of standard algebraic Riccati equations and linear matrix inequalities,

    F. J. Vargas and R. A. Gonzalez, “On the existence of a stabilizing solution of modified algebraic Riccati equations in terms of standard algebraic Riccati equations and linear matrix inequalities,” IEEE Control Systems Letters, vol. 4, no. 1, pp. 91–96, 2020

    2020

  20. [27]

    Comments “On pole assignment in multi-input controllable linear systems

    M. Heymann and W. Wonham, “Comments “On pole assignment in multi-input controllable linear systems”,” IEEE Transactions on Auto- matic Control, vol. 13, no. 6, pp. 748–749, 1968

    1968

  21. [29]

    Active fault-tolerant control for an internet-based networked three-tank system,

    X. He, Z. Wang, L. Qin, and D. Zhou, “Active fault-tolerant control for an internet-based networked three-tank system,” IEEE Transactions on Control Systems Technology, vol. 24, no. 6, pp. 2150–2157, 2016

    2016

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.