Detecting Switching Attacks On Traffic Flow Regulation For Changing Driving Patterns

Sanchita Ghosh; Tanushree Roy

REVIEW 3 major objections 5 minor 32 references

Detecting Switching Attacks On Traffic Flow Regulation For Changing Driving Patterns

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A bank of detectors can catch corrupted ramp-meter switching on freeways.

desk verdict A plausible detector-bank scheme for ramp-metering switching attacks, but the sensitivity guarantee rests on an invalid min/integral step and the gains require the unknown mode; the formal claims do not stand as written. read the letter →

arxiv 2505.23033 v1 pith:NHH62RJR submitted 2025-05-29 eess.SY cs.SY

Sanchita Ghosh , Tanushree Roy This is my paper

classification eess.SYcs.SY

keywords switchingattacksrampmeteringmulti-modaltrafficAw-Rascle-Zhangmodelbacksteppingresidualgenerationcyber-physicalsystemsLMIdesign

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 a practical way to detect cyberattacks that corrupt the switching command of a ramp-metering controller when traffic can run in several modes, such as rainy, light, or heavy traffic. The central claim is that a bank of detectors, one for each admissible traffic mode, can produce a residual signal that stays small during legitimate mode changes and grows when the switching command is attacked. The paper derives formal performance guarantees for this residual: exponential stability under normal operation, boundedness under uncertainties and attacks, robustness to uncertainties, and sensitivity to switching attacks. If correct, this would let a traffic supervisor delay mode identification without raising false alarms, while still catching denial-of-service and false-data-injection attacks on the switching command.

What carries the argument

The load-bearing mechanism is a bank of m detectors (11)-(14), one for each admissible traffic mode, whose outputs ζ^j are combined into the residual r(t)=min_j |ζ^j(t)|. Backstepping transformations (30)-(31) and their inverses (37)-(38) decouple the coupled error dynamics, and a family of Lyapunov functions (50)-(51) is used to derive the LMI-based gain conditions in Theorem 1. The residual is compared against a threshold J set by a target false-alarm probability, and an attack is declared when the residual crosses this threshold.

What would settle it

Simulate the nonlinear ARZ model (1)-(3) with a rain-induced switch from mode 2 to mode 1, with no attack, and let the state pass through a realistic transient far from the new steady state; if the residual r(t)=min_j |ζ^j(t)| crosses the designed threshold J during nominal switching, the claimed robustness guarantee fails on the actual nonlinear dynamics. The direct calculation to check is whether min_j Θ^j=0 holds along a non-steady-state trajectory in the sense required by equation (71).

Watch

Extended reading notes

Core claim

The paper claims that for a freeway segment modeled by the Aw-Rascle-Zhang traffic PDE with m possible traffic modes, the detection residual r(t)=min_j |ζ^j(t)|, defined as the smallest output among m output-injection detectors, can distinguish legitimate controller switching from corrupted switching. The main theorem states that if there exists a free parameter ξ and detector gains satisfying the linear matrix inequalities (44)-(46), then the residual is exponentially stable under nominal operation, bounded under uncertainties and attacks in the sense of anomaly/uncertainty-to-residual stability, robust to uncertainties, and sensitive to switching attacks. The scheme is demonstrated on two attack scenarios: a denial-of-service attack that prevents the ramp meter from switching to the rainy-weather mode, detected within 8 seconds, and a false-data-injection attack that switches the meter to a lighter mode, detected within 50 seconds.

Load-bearing premise

The analysis assumes that during a mode change the traffic state is always exactly described by one of the m linearized ARZ modes around its steady state, so the matching detector has zero parameter mismatch and unknown transient and linearization errors are absent.

Editorial extensions

If this is right

A supervisory controller can take time to identify a changed traffic mode without generating false alarms, because the detector matching the true mode keeps the minimum residual small.
Denial-of-service and false-data-injection attacks on the ramp-meter switching command are detectable in real time, within seconds in the simulated scenarios.
The LMI conditions in Theorem 1 provide a constructive procedure for choosing detector gains that satisfy formal stability, robustness, and sensitivity criteria.
The bank-of-detectors residual logic could be applied to other switched distributed-parameter systems where the operating mode is uncertain and mode identification is delayed.

Reading between the lines

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

The strongest untested assumption is that during a mode change the traffic state is always exactly one of the linearized steady-state modes, so that the matching detector has zero parameter mismatch; evaluating the detector bank on the full nonlinear ARZ model with realistic transients would test whether this assumption is safe.
The threshold J is defined through a nominal residual distribution, but in practice that distribution must be estimated online; the detection delay and false-alarm rate may depend on how well that estimate matches real traffic noise.
The min-over-detectors residual could hide an attack that occurs during a mode transition, since the newly matched detector's transient may also push the residual up; separating transient switching effects from attacks is an open practical issue.
A testable extension is to replace the steady-state modes with a continuum of fundamental-diagram curves and check whether the finite detector bank still yields a small matching residual under gradual weather changes.

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.

Desk Editor's Note

read the letter

Quick take: the paper applies the authors' detector-bank idea to multimodal ARZ ramp metering, and the simulations look reasonable. But the main theorem's sensitivity guarantee is not proven. The proof derives a lower bound for each detector's output energy, then "takes the minimum" to get a lower bound on the residual energy. That step is invalid: r is the minimum of |ζ^j|, so ∫r² is ≤ each ∫ζ^{j,2}, not ≥. A lower bound on every detector does not lower-bound the minimum. The matched detector may have a large output while a mismatched one stays quiet, so the min residual can remain below threshold. This breaks the missed-detection guarantee (25) and with it the central detection claim.

There is a second issue that is just as serious: the detector gains in (35) are expressed using α, the unknown true mode. The backstepping kernels (33)-(34) and gains depend on the very quantity the bank is meant to identify. You could redesign the gains per detector using that detector's own mode, but that is not what the paper does; the error dynamics are coupled through the true plant parameters, so the transformation as written requires knowing α. That makes the design circular.

To give credit: the problem is well motivated, and the detector-bank structure is a reasonable way to handle mode uncertainty if the gains can be made mode-independent or adaptive. The LMI formulation is concrete, and the DoS/FDI simulations show the residual does rise after attacks. The robustness and AURS parts look more defensible—there the min step on upper bounds is valid, though I would want to check the details. The heavy reliance on the authors' prior backstepping results is fine as long as those results are solid.

Bottom line: this is not a desk-reject mess. The concept is plausible and the paper is carefully written. But the two load-bearing flaws mean the formal guarantees are not established. If the authors can fix the sensitivity argument (e.g., redefine the residual as a sum or max, or prove a lower bound on the min directly) and remove the α-dependence in the gains, it could become a solid niche contribution. As it stands, I would not accept it. I would still send it to review: a good referee can identify these issues, and the paper deserves the community's time.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes a bank of m output-injection detectors for a multimodal Aw-Rascle-Zhang freeway model, with the goal of detecting switching attacks on ramp-metering commands when the traffic mode is uncertain. The residual is the minimum of the detector outputs, and the authors derive LMI-based conditions (Theorem 1) intended to guarantee exponential stability, anomaly/uncertainty-to-residual stability, robustness, and sensitivity. The paper also presents simulation case studies for denial-of-service and false-data-injection attacks under a three-mode traffic scenario.

Significance. If the theoretical guarantees were correct, the paper would provide a valuable analytical framework for attack detection in hierarchical ramp-metering under mode uncertainty, and the combination of a detector bank with backstepping-based residual generation is a sensible approach. The four-criteria formulation (ES, AURS, robustness, sensitivity) is useful, and the simulation scenarios are realistic. However, the central sensitivity proof is invalid, and the detector-gain construction appears to depend on the unknown mode, so the main claims are not established. The paper does not provide code or machine-checked proofs, and the analytical proof as written is difficult to verify independently.

major comments (3)

[Appendix, Condition for sensitivity, Eq. (79) to Eq. (25)] The step 'taking minimum on both sides of (79)' is invalid for the chosen residual. From (79), for every j one obtains ∫_0^∞ ζ^{j2} dt ≥ Υ6 ∫_0^∞ (δ^2 + Θ^j) dt − ε, so taking the minimum over j gives only min_j ∫_0^∞ ζ^{j2} dt ≥ Υ6 ∫_0^∞ δ^2 dt − ε, even granting min_j Θ^j = 0. But by (20), r(t) = min_j |ζ^j(t)|, hence r^2(t) = min_j ζ^{j2}(t), and therefore ∫_0^∞ r^2 dt = ∫_0^∞ min_j ζ^{j2}(t) dt ≤ min_j ∫_0^∞ ζ^{j2}(t) dt. The inequality points in the wrong direction: a lower bound on the minimum of the integrals does not lower-bound the integral of the minimum. Consequently, the sensitivity criterion (25), and with it the missed-detection guarantee of Theorem 1, is not proven by the argument given. This is a load-bearing gap, not a presentation issue: the proof would need a different mechanism to establish a positive lower bound on r directly.
[Section 3.3, Eqs. (33)–(35)] The kernels R^j, S^j and the detector gains k^j_1, k^j_2 are written using the true-mode parameters v^α_*, h^α, γp^α_*, and c^α, but α is precisely the unknown mode that the detector bank is intended to compensate. If the formulas are literal, the gains for the j-th detector cannot be computed without knowing α, which defeats the purpose of the bank. If the intention was to use j in place of α, then the backstepping cancellation is not established, because the error dynamics (15)–(18) contain α-dependent coefficients. Either reading leaves a load-bearing gap in the design.
[Section 3.4, Eq. (71)] The proof relies on min_j Θ^j = 0, which requires that at every time the true traffic state is exactly described by one of the linearized modes with zero parameter mismatch. The paper provides no bound on the linearization error or on the transient that occurs during a legitimate mode change when the state is not at the new steady state. Since the motivating scenario is precisely changing driving patterns, this assumption is not innocuous: otherwise the matched detector may not be matched, and nominal switching could produce a residual that triggers false alarms.

minor comments (5)

[Section 3.4, Condition for ES] The paragraph 'Condition for ES' states 'δ, ηq, ηv ≠ 0' where the nominal no-uncertainty condition should be δ = ηq = ηv = 0; this typo makes the proof harder to follow.
[Section 2.1] In the sentence introducing the uncertainties, 'ηq, ηq ∈ R' should presumably read 'ηq, ηv ∈ R'.
[Table 1 and Appendix proof] Several parameters in the proof are undefined or inconsistent: for example, the first row of Table 1 for Υ^{α,j}_8 is garbled, µ17–µ25 are introduced in the proof without complete definitions, and µ20 is reused with different meanings in different parts of the proof. This prevents independent verification of the LMI conditions.
[Section 4, threshold definition] The threshold is stated to be 0.02 'obtained using (21)', but the distribution P(rη) and the procedure for computing J are not described, so the threshold selection is not reproducible.
[Section 4, sentence] The sentence 'we consider have 3 admissible traffic modes' contains a typo and should be reworded.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the detector design and stability analysis are derived from the PDE model and Lyapunov arguments, not from fitted outputs.

full rationale

The paper's central claim is a model-based detector bank with LMI design conditions, and the derivation chain does not reduce to its own inputs. The residual r(t)=min_j |\zeta^j(t)| is a definition, not a fitted quantity, and the detector gains k1^j,k2^j,k3^j are designed from backstepping transformations and Lyapunov inequalities rather than calibrated to the detection outcomes. The use of the authors' prior work [3,5,32] is methodological: it supplies the backstepping framework, the detector-bank concept, and the performance-criteria definitions, while the present proof supplies its own Lyapunov estimates and LMI conditions. No load-bearing conclusion is justified solely by a self-citation, and no uniqueness theorem or ansatz is imported to forbid alternatives. The proof does contain a serious non-circular gap in the sensitivity part, where a lower bound for each detector output is incorrectly taken to lower-bound the minimum residual; that is a correctness or mathematical-validity concern, not a circularity, because it does not make the claimed result equivalent to an input by construction. The paper is therefore self-contained with respect to circularity, and the honest finding is no significant circularity.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests on a linearized PDE model, exact knowledge of all mode parameters for the detector bank, bounded uncertainties, and an unproven backstepping transformation. The design also introduces several free tuning parameters and a hand-picked threshold. No new physical entities are postulated.

free parameters (5)

Detection threshold J = 0.02
Chosen for each mode with no reported distribution or calibration data; affects false alarm and detection performance.
LMI tuning parameters μ1..μ16 = not reported
Free positive constants in Theorem 1 that shape the detector gains; no numerical values or feasibility solutions are given.
Mode parameters in Table 2 = v_f, ρ, v_*, q_*, k_σ per mode
Used in simulations; physically motivated but chosen by the authors, and the detector design assumes they are known exactly.
Uncertainty and noise magnitudes = 0.12%, ±2.5 m/s, ±10%, ±2%
Chosen for the case studies to imitate real-world conditions; no sensitivity analysis is reported.
Detector gains k^j_1, k^j_2, k^j_3
Design variables constrained by LMIs; numerical values are not reported, and formulas in (35) depend on the true mode.

assumptions (6)

domain assumption ARZ PDE (1)-(3) is an accurate model of freeway traffic
The entire detector design is based on this model and its linearization.
domain assumption The traffic state is always near the mode-α steady state so the linearized system (7)-(10) is valid
The analysis uses the linearized W,V dynamics even during mode transitions and attacks, without bounding linearization error.
domain assumption The true mode α is one of the m known modes and the corresponding detector parameters match it exactly
Eq. (71) sets min_j Θ^j=0, which requires a perfect parameter match for one detector.
domain assumption Uncertainties η_q, η_v are bounded and square-integrable
The L2 gains in (23)-(25) require integrable disturbances; this is not explicitly stated.
ad hoc to paper Initial conditions of the backstepped system are bounded as in Assumption 1
Explicit assumption in §3.4 used to bound r(0) and the Lyapunov functions.
standard math The backstepping transformation (26)-(29) with kernels (33)-(34) exists and is invertible
Stated as standard methodology [3] with the detailed derivation omitted, so the paper relies on prior theory without proof.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Detecting Switching Attacks On Traffic Flow Regulation For Changing Driving Patterns." pith.science (2026). https://pith.science/paper/NHH62RJR

@misc{pith2026250523033,
  author       = {Pith},
  title        = {Pith review of: Detecting Switching Attacks On Traffic Flow Regulation For Changing Driving Patterns},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NHH62RJR}},
  note         = {Machine review of arXiv:2505.23033}
}

read the original abstract

Modern traffic management systems increasingly adopt hierarchical control strategies for improved efficiency and scalability, where a local traffic controller mode is chosen by a supervisory controller based on the changing large-scale driving patterns. Unfortunately, such local metering controllers are also vulnerable to cyberattacks that can disrupt the controller switching, leading to undesired, inefficient, and even unsafe traffic operations. Additionally, the detection of such attacks becomes challenging when the operational mode of the traffic is uncertain and the operational mode identification is delayed. Thus, in this work, we propose a cyberattack detection scheme to detect the compromised controller switching in ramp metering for an uncertain, multimodal macroscopic traffic operation of a freeway segment. In particular, we propose a bank of detectors corresponding to each admissible traffic mode that can compensate for the uncertain traffic mode of the freeway. Furthermore, we utilize backstepping tools along with Lyapunov function theory to achieve analytical performance guarantees for the detector, such as nominal exponential stability, anomaly/uncertainty-to-residual stability, robustness, and sensitivity. Finally, we demonstrate the efficacy of the proposed detection scheme through simulations of free traffic under realistic traffic parameters, uncertainties, and commonly occurring attack scenarios.

Figures

Figures reproduced from arXiv: 2505.23033 by the authors.

**Figure 2.** Figure depicts the impact of controller switching on traffic behavior under nominal conditions and DoS [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Under DoS attack, the figure shows traffic density (top), flux (middle), and velocity (bottom). [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Under DoS attack, the figure shows nominal and corrupted boundary ramp control (top) and attack detection [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

**Figure 5.** Figure 5: Under FDI attack, the figure shows traffic density (top), flux (middle), and velocity (bottom). [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

**Figure 6.** Figure 6: Under FDI attack, the figure shows nominal and corrupted boundary ramp control (top) and attack detection [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

[1]

Smart Cities As Cyber-physical Social Systems

Christos G Cassandras. Smart Cities As Cyber-physical Social Systems. Engineering, 2(2):156–158, 2016

work page 2016
[2]

Implications Of Traffic Signal Cybersecurity On Potential Deliberate Traffic Disruptions

Kenneth A Perrine, Michael W Levin, Cesar N Yahia, Melissa Duell, and Stephen D Boyles. Implications Of Traffic Signal Cybersecurity On Potential Deliberate Traffic Disruptions. Transportation research part A: policy and practice, 120:58–70, 2019

work page 2019
[3]

Secure traffic networks in smart cities: Analysis and design of cyber-attack detection algorithms

Tanushree Roy and Satadru Dey. Secure traffic networks in smart cities: Analysis and design of cyber-attack detection algorithms. In 2020 American Control Conference (ACC), pages 4102–4107. IEEE, 2020

work page 2020
[4]

A socio-technical approach for resilient connected transportation systems in smart cities

Tanushree Roy, Amara Tariq, and Satadru Dey. A socio-technical approach for resilient connected transportation systems in smart cities. IEEE Transactions on Intelligent Transportation Systems, 23(6):5019–5028, 2021

work page 2021
[5]

Security of cyber-physical systems under compromised switching

Sanchita Ghosh and Tanushree Roy. Security of cyber-physical systems under compromised switching. In 2023 IEEE Conference on Control Technology and Applications (CCTA), pages 1034–1039. IEEE, 2023

work page 2023
[6]

Comparative model accuracy of a data-fitted generalized Aw-Rascle-Zhang model

Shimao Fan, Michael Herty, and Benjamin Seibold. Comparative Model Accuracy Of A Data-fitted Generalized Aw-Rascle-Zhang Model. arXiv preprint arXiv:1310.8219, 2013

work page Pith review arXiv 2013
[7]

An Extended Kalman Filter Application For Traffic State Estimation Using CTM With Implicit Mode Switching And Dynamic Parameters

Chris MJ Tamp `ere and LH Immers. An Extended Kalman Filter Application For Traffic State Estimation Using CTM With Implicit Mode Switching And Dynamic Parameters. In2007 IEEE Intelligent Transportation Systems Conference, pages 209–216. IEEE, 2007

work page 2007
[8]

Distributionally Robust Ramp Metering Under Traffic Demand Uncertainty

Chuanye Gu, Changzhi Wu, Yonghong Wu, and Benchawan Wiwatanapataphee. Distributionally Robust Ramp Metering Under Traffic Demand Uncertainty. Transportmetrica B: Transport Dynamics, 10(1):652–666, 2022

work page 2022

Show all 32 references

[9]

Urban Traffic Signal Control Robust Optimization Against Risk-averse And Worst-case Cyberattacks

Liang Zheng, Ji Bao, and Zhenyu Mei. Urban Traffic Signal Control Robust Optimization Against Risk-averse And Worst-case Cyberattacks. Information Sciences, 640:119067, 2023

2023
[10]

Distributed Cooperative Coverage Control Of Sensor Networks

Wei Li and Christos G Cassandras. Distributed Cooperative Coverage Control Of Sensor Networks. In Proceed- ings Of The 44Th IEEE Conference On Decision And Control, pages 2542–2547. IEEE, 2005

2005
[11]

Two-layer Adaptive Signal Control Frame- work For Large-scale Dynamically-congested Networks: Combining Efficient Max Pressure With Perimeter Control

Dimitrios Tsitsokas, Anastasios Kouvelas, and Nikolas Geroliminis. Two-layer Adaptive Signal Control Frame- work For Large-scale Dynamically-congested Networks: Combining Efficient Max Pressure With Perimeter Control. Transportation Research Part C: Emerging Technologies, 152:...

2023
[12]

Ramp Metering Strategies: A Literature Review

Patricia Cazorla, Francisco Calder ´on, and Elina Avila-Ord´o˜nez. Ramp Metering Strategies: A Literature Review. Revista Polit´ecnica, 50(1):15–26, 2022

2022
[13]

Creating Complex Congestion Pat- terns Via Multi-objective Optimal Freeway Traffic Control With Application To Cyber-security

Jack Reilly, S ´ebastien Martin, Mathias Payer, and Alexandre M Bayen. Creating Complex Congestion Pat- terns Via Multi-objective Optimal Freeway Traffic Control With Application To Cyber-security. Transportation Research Part B: Methodological, 91:366–382, 2016

2016
[14]

Green Lights Forever: Analyzing The Security Of Traffic Infrastructure

Branden Ghena, William Beyer, Allen Hillaker, Jonathan Pevarnek, and J Alex Halderman. Green Lights Forever: Analyzing The Security Of Traffic Infrastructure. In 8Th USENIX Workshop On Offensive Technologies (WOOT 14), 2014

2014
[15]

Effects Of Cyberattacks On Regional Traffic Networks In A Connected Vehicle Environment.IEEE Transactions on Network Science and Engineering, 2025

Liangwen Wang, Heng Ding, Xiaoyan Zheng, and Weihua Zhang. Effects Of Cyberattacks On Regional Traffic Networks In A Connected Vehicle Environment.IEEE Transactions on Network Science and Engineering, 2025

2025
[16]

Eval- uating Impact Of Remote-access Cyber-attack On Lane Changes For Connected Automated Vehicles

Changyin Dong, Yujia Chen, Hao Wang, Leizhen Wang, Ye Li, Daiheng Ni, De Zhao, and Xuedong Hua. Eval- uating Impact Of Remote-access Cyber-attack On Lane Changes For Connected Automated Vehicles. Digital Communications and Networks, 10(5):1480–1492, 2024

2024
[17]

A Cyberattack Detection-isolation Algorithm For CA V Under Changing Driving Environment

Sanchita Ghosh, Nutan Saha, and Tanushree Roy. A Cyberattack Detection-isolation Algorithm For CA V Under Changing Driving Environment. IEEE Transactions on Intelligent Transportation Systems, 2024

2024
[18]

Assessment Of Cyberattack Detection-Isolation Algorithm For CA V Pla- toons Using SUMO

Sanchita Ghosh and Tanushree Roy. Assessment Of Cyberattack Detection-Isolation Algorithm For CA V Pla- toons Using SUMO. arXiv preprint arXiv:2503.14628, 2025

2025
[19]

Cyber-attack detection in socio-technical transportation systems exploiting redundancies between physical and social data

Tanushree Roy, Sara Sattarzadeh, and Satadru Dey. Cyber-attack detection in socio-technical transportation systems exploiting redundancies between physical and social data. IEEE Transactions on Systems, Man, and Cybernetics: Systems, pages 1–12, 2023

2023
[20]

Stealthy Attack Detection Of Controlled Ramp Meters In Freeway Networks

Charalambos Menelaou, Kangkang Zhang, Stelios Timotheou, Christos G Panayiotou, and Thomas Parisini. Stealthy Attack Detection Of Controlled Ramp Meters In Freeway Networks. In 2024 IEEE 63Rd Conference On Decision And Control (CDC), pages 1289–1294. IEEE, 2024

2024
[21]

Spoofing Cyber Attack Detection In Probe-based Traffic Moni- toring Systems Using Mixed Integer Linear Programming

Edward S Canepa and Christian G Claudel. Spoofing Cyber Attack Detection In Probe-based Traffic Moni- toring Systems Using Mixed Integer Linear Programming. In 2013 International Conference On Computing, Networking And Communications (ICNC), pages 327–333. IEEE, 2013

2013
[22]

Detection Of Cyber-attacks In Automotive Traffic Using Macroscopic Models And Gaussian Processes

Abhishek Kashyap, Animesh Chakravarthy, and Prathyush P Menon. Detection Of Cyber-attacks In Automotive Traffic Using Macroscopic Models And Gaussian Processes. IEEE Control Systems Letters, 6:1688–1693, 2021

2021
[23]

Vulnerability Mitigation Of Urban Traffic Control Against Cyberattacks Using Secure Multi-Party Computation

Saeed Adelipour, Enayatollah Amiri Darreh Razgahi, and Mohammad Haeri. Vulnerability Mitigation Of Urban Traffic Control Against Cyberattacks Using Secure Multi-Party Computation. IEEE Transactions on Intelligent Transportation Systems, 2025

2025
[24]

A Physics-informed Deep Learning Paradigm For Traffic State And Fundamental Diagram Estimation

Rongye Shi, Zhaobin Mo, Kuang Huang, Xuan Di, and Qiang Du. A Physics-informed Deep Learning Paradigm For Traffic State And Fundamental Diagram Estimation. IEEE Transactions on Intelligent Transportation Sys- tems, 23(8):11688–11698, 2021

2021
[25]

Aw and Michel Rascle

A.A.T.M. Aw and Michel Rascle. Resurrection Of” Second Order” Models Of Traffic Flow. SIAM journal on applied mathematics, 60(3):916–938, 2000

2000
[26]

J. P. Hespanha, Daniel Liberzon, and A. Stephen Morse. Hysteresis-based switching algorithms for supervisory control of uncertain systems. Automatica, 39:263–272, 2 2003

2003
[27]

Mode-dependent Spatiotemporal Event-triggered Control For Switched Nonlinear PDE Systems With Persistent Dwell-time Switching Regulation

Xiaona Song, Zenglong Peng, Shuai Song, Xin Wang, and Xiaohui Zhang. Mode-dependent Spatiotemporal Event-triggered Control For Switched Nonlinear PDE Systems With Persistent Dwell-time Switching Regulation. Journal of Control and Decision, pages 1–17, 2025

2025
[28]

Traffic Congestion Control For Aw–Rascle–Zhang Model

Huan Yu and Miroslav Krstic. Traffic Congestion Control For Aw–Rascle–Zhang Model. Automatica, 100:38– 51, 2019

2019
[29]

Constructing A Fundamental Diagram For Traffic Flow With Automated Vehicles: Methodology And Demonstration

Xiaowei Shi and Xiaopeng Li. Constructing A Fundamental Diagram For Traffic Flow With Automated Vehicles: Methodology And Demonstration. Transportation Research Part B: Methodological, 150:279–292, 2021

2021
[30]

An Empirical Study On Parameters Affecting Traffic Stream Variables Under Rainy Conditions

Archana Nigam, Manish Chaturvedi, and Sanjay Srivastava. An Empirical Study On Parameters Affecting Traffic Stream Variables Under Rainy Conditions. In2022 14Th International Conference On COMmunication Systems & NETworkS (COMSNETS), pages 818–823. IEEE, 2022

2022
[31]

Model-based Fault Diagnosis Techniques: Design Schemes, Algorithms, And Tools

Steven X Ding. Model-based Fault Diagnosis Techniques: Design Schemes, Algorithms, And Tools . Springer Science & Business Media, 2008

2008
[32]

Actuator anomaly detection in linear parabolic distributed parameter cyber- physical systems

Tanushree Roy and Satadru Dey. Actuator anomaly detection in linear parabolic distributed parameter cyber- physical systems. IEEE Transactions on Control Systems Technology, pages 1–12, 2023

2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Smart Cities As Cyber-physical Social Systems

Christos G Cassandras. Smart Cities As Cyber-physical Social Systems. Engineering, 2(2):156–158, 2016

work page 2016

[2] [2]

Implications Of Traffic Signal Cybersecurity On Potential Deliberate Traffic Disruptions

Kenneth A Perrine, Michael W Levin, Cesar N Yahia, Melissa Duell, and Stephen D Boyles. Implications Of Traffic Signal Cybersecurity On Potential Deliberate Traffic Disruptions. Transportation research part A: policy and practice, 120:58–70, 2019

work page 2019

[3] [3]

Secure traffic networks in smart cities: Analysis and design of cyber-attack detection algorithms

Tanushree Roy and Satadru Dey. Secure traffic networks in smart cities: Analysis and design of cyber-attack detection algorithms. In 2020 American Control Conference (ACC), pages 4102–4107. IEEE, 2020

work page 2020

[4] [4]

A socio-technical approach for resilient connected transportation systems in smart cities

Tanushree Roy, Amara Tariq, and Satadru Dey. A socio-technical approach for resilient connected transportation systems in smart cities. IEEE Transactions on Intelligent Transportation Systems, 23(6):5019–5028, 2021

work page 2021

[5] [5]

Security of cyber-physical systems under compromised switching

Sanchita Ghosh and Tanushree Roy. Security of cyber-physical systems under compromised switching. In 2023 IEEE Conference on Control Technology and Applications (CCTA), pages 1034–1039. IEEE, 2023

work page 2023

[6] [6]

Comparative model accuracy of a data-fitted generalized Aw-Rascle-Zhang model

Shimao Fan, Michael Herty, and Benjamin Seibold. Comparative Model Accuracy Of A Data-fitted Generalized Aw-Rascle-Zhang Model. arXiv preprint arXiv:1310.8219, 2013

work page Pith review arXiv 2013

[7] [7]

An Extended Kalman Filter Application For Traffic State Estimation Using CTM With Implicit Mode Switching And Dynamic Parameters

Chris MJ Tamp `ere and LH Immers. An Extended Kalman Filter Application For Traffic State Estimation Using CTM With Implicit Mode Switching And Dynamic Parameters. In2007 IEEE Intelligent Transportation Systems Conference, pages 209–216. IEEE, 2007

work page 2007

[8] [8]

Distributionally Robust Ramp Metering Under Traffic Demand Uncertainty

Chuanye Gu, Changzhi Wu, Yonghong Wu, and Benchawan Wiwatanapataphee. Distributionally Robust Ramp Metering Under Traffic Demand Uncertainty. Transportmetrica B: Transport Dynamics, 10(1):652–666, 2022

work page 2022