Pith. sign in

REVIEW 3 major objections 6 minor 36 references

Tunable Leg Stiffness in a Monopedal Hopper for Energy-Efficient Vertical Hopping Across Varying Ground Profiles

T0 review · 3 major / 6 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 A hopping robot with real-time tunable leg stiffness can find, for each combination of ground stiffness and damping, a stiffness setting that maximizes steady-state hopping height at fixed energy input.

desk verdict A promising tunable-stiffness hopper with a clean problem formulation, but the submitted text stops before the results and the constant-energy definition is too vague to support the headline claim. read the letter →

arxiv 2508.02873 v2 pith:EMTSB3N3 submitted 2025-08-04 cs.RO cs.SYeess.SY

Rongqian Chen , Jun Kwon , Kefan Wu , Wei-Hsi Chen This is my paper · ORCID
classification cs.ROcs.SYeess.SY
keywords tunablelegstiffnessverticalhoppingenergy-efficientlocomotiongrounddampingmonopedalrobotpneumaticbellowsactuatorapexheight
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 asks whether a hopping robot can save energy by adjusting leg stiffness to the ground it lands on. It presents HASTA (Hopper with Adjustable Stiffness for Terrain Adaption), a vertical hopper whose pneumatic bellows leg can change stiffness in real time, and tests it against a ground emulator that sets both ground stiffness and damping. The central result is that for every tested combination of ground stiffness and damping, one leg-stiffness setting from the robot's range gives the highest steady-state hopping height at a fixed energy input per hop. That optimum follows a trend: soft, damped ground favors softer legs, while hard, lightly damped ground favors stiffer legs. The paper argues that steady-state apex height is a valid proxy for energy-efficient hopping.

What carries the argument

The central object is HASTA's tunable-stiffness leg, built from three pneumatic bellows actuators that change leg stiffness by a factor of 1.43 without an external air supply, plus a tendon-driven motor that pre-compresses the leg spring to store a fixed input energy per hop. The argument runs through the steady-state apex hopping height, defined as the maximum vertical displacement after hopping height stabilizes, with the ground modeled as a parallel spring-damper whose stiffness and damping are programmed by a five-bar-linkage ground emulator. A lookup table maps motor angle and leg stiffness to leg compression, and the experiment searches the discrete leg-stiffness set for the setting that maximizes apex height on each ground profile. The simulation reproduces the experiment with a linear mass-spring-damper model and then extends the search across finer grids of ground stiffness, damping, leg damping, and energy input.

What would settle it

Instrument HASTA to log motor current and bellows pressure over complete hop cycles on a single ground profile, compute actual energy consumed per hop for each leg stiffness, and compare the stiffness that maximizes apex height with the stiffness that minimizes energy per meter of height; a mismatch would falsify the claim that apex height is the energy-efficiency metric.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is an experimentally repeatable, ground-dependent optimum in leg stiffness for vertical hopping. With three leg stiffnesses (3351, 4279, and 5341 N/m) and a grid of ground stiffness and damping values, HASTA reaches its maximum steady-state apex height at a different optimal leg stiffness depending on the ground profile. As ground stiffness increases or damping decreases, stiffer legs reduce energy loss and achieve higher hopping heights; softer legs are better on soft, damped ground because they minimize penetration and energy loss. The same trend appears in a simplified mass-spring-damper simulation, which the paper takes as evidence that a controller could select leg stiffness from ground-property estimates.

Load-bearing premise

The load-bearing premise is that steady-state apex hopping height is a faithful proxy for energy-efficient hopping: if apex height does not capture losses that vary with stiffness, such as motor heating, pneumatic losses, or stance-phase dissipation, then the stiffness that maximizes height may not be the one that minimizes energy per hop.

Editorial extensions

If this is right

  • A robot that can change leg stiffness in real time will out-hop a fixed-stiffness robot on the same ground at the same energy input.
  • Optimal stiffness is ground-dependent: soft, damped ground calls for softer legs, and hard, lightly damped ground calls for stiffer legs.
  • The simplified simulation reproduces the direction of the experimental effect, so it can precompute stiffness choices for ground profiles not yet tested.
  • Ground damping matters as much as ground stiffness in choosing leg stiffness; selecting stiffness from ground stiffness alone would leave energy savings unused.

Reading between the lines

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

  • A natural extension the paper leaves open is closing the loop online: estimate ground stiffness and damping from touchdown transients and servo the bellows to the predicted optimal stiffness within a few hops.
  • The same stiffness-selection logic should transfer to horizontal running, where leg stiffness also trades off against speed and stance time; the hopper's discrete optimal-stiffness map gives a baseline prediction for those gaits.
  • The paper's trend predicts an extreme-case test: on the softest, most damped ground the softest leg setting should dominate, and on the hardest, least damped ground the stiffest setting should dominate.
  • Because the simulation used a simplified mass-spring-damper, a mismatch between simulated and experimental optimal stiffness on untested ground profiles would reveal which unmodeled effects, such as toe contact or actuator dynamics, matter most.
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

3 major / 6 minor

Summary. The paper presents HASTA, a vertically constrained monopedal hopper with in-situ tunable pneumatic leg stiffness, and studies how leg stiffness should be selected for energy-efficient vertical hopping on grounds with different stiffness and damping. The central claim is that, at a fixed input energy per hop, there is a per-ground-profile leg stiffness that maximizes the steady-state apex height, and that this optimal stiffness tends to be larger on stiff, lightly damped ground and smaller on soft, highly damped ground. The authors also propose a simplified mass-spring-damper simulation to validate the experimental observations. The provided manuscript text includes the introduction, problem statement, and the beginning of the experimental section, but ends before the results section (Section V) and the simulation-validation comparison are presented.

Significance. If the full results support the claims, the paper would be a valuable experimental demonstration of hardware-based tunable stiffness improving a locomotion energy metric across a range of ground profiles, including damping, which prior work has largely neglected. The design description is detailed enough to be reproducible, and the use of a ground emulator with programmable stiffness and damping is a strong methodological choice. However, the absence of the results section in the provided text means the central empirical claim cannot currently be assessed, and the energy-efficiency interpretation has not yet been grounded in any measured energy data.

major comments (3)
  1. [Section V / Contribution 3] The provided manuscript ends in Section III.B, before the results (Section V) and the simulation-experiment comparison that would substantiate contribution 3. The abstract and introduction assert that the best leg stiffness was found experimentally and that the simulation "demonstrates similar behavior," but no figures, tables, error bars, or statistical analyses are present in the text under review. The central claim of the paper is therefore unsupported by the submitted material. This is a load-bearing omission: the paper cannot be evaluated until the empirical results, including the steady-state apex heights for each leg-stiffness and ground-profile combination and the simulation-validation data, are included.
  2. [Section III.A and Table I] The problem statement in Section II defines a constant input potential energy E_in per hop, and Table I lists E_in = 0.97 J, but the experimental protocol never specifies how leg pre-compression δ is chosen for each stiffness setting. The text states that a lookup table δ = f(θ_m, k_l) was generated experimentally, but it does not state that δ is selected so that (1/2)k_l δ² = E_in for every k_l. If a single δ were used across all stiffness values, stiffer legs would store more spring energy and the observed height differences would be an artifact of unequal input energy rather than evidence about stiffness adaptation. The authors must clearly state the pre-compression protocol and, ideally, report the measured δ and computed stored energy for each tested stiffness.
  3. [Section II] The paper equates steady-state apex hopping height with energy efficiency: Section II says "We argue that this steady-state hopping height is directly related to energy-efficient hopping." No direct energy measurement is reported; the ESP32 records current and voltage, but no electrical-energy-per-hop result appears in the provided text. If motor losses, pneumatic losses, or stance-phase dissipation vary with leg stiffness, maximizing apex height at constant mechanical input energy does not necessarily minimize total energy consumption per hop. The authors should either report measured energy consumption for each stiffness setting or provide a quantitative model showing that apex height is a valid surrogate for energy efficiency across the stiffness range.
minor comments (6)
  1. [Abstract] There is a typo in the abstract: "Adap-tion" should be "Adaptation."
  2. [Section II and Table I] Many mathematical symbols are garbled in the rendered text (e.g., mass variables, stiffness symbols, and the energy expression in Problem 1). Please ensure the PDF renders all equations and symbols legibly.
  3. [Section III.A] The lookup table relating motor angle, leg stiffness, and pre-compression is described only as "experimentally generated." A short description of the calibration procedure, number of samples, and repeatability would help the reader judge the accuracy of the reported compression control.
  4. [Table I] The simulation leg damping coefficients (30, 35, 40 N·s/m) do not overlap with the experimental ground damping values (17.1 to 71.4 N·s/m), and the experimental leg damping is not listed. The authors should justify the simulation damping values and, if possible, identify the experimental leg damping coefficient.
  5. [References] References [26]–[35] appear unrelated to the topic of tunable-stiffness hopping (they concern network scheduling, federated reinforcement learning, and communication systems). The authors should either cite these works in context or remove them to avoid distracting from the relevant literature.
  6. [Figure 2 caption] The caption refers to "yellow and green text boxes" to distinguish simulation and experimental guards, but the figure is grayscale in the provided version. Use symbols or line styles that are legible in grayscale.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the optimal stiffness map is an experimental characterization, not a derivation from its own output.

full rationale

The paper's central claim is an experimental characterization: for each ground profile, the leg stiffness from a discrete set that maximizes steady-state apex height at fixed input energy is found by direct measurement on the HASTA robot with a programmable ground emulator, not derived from a model that already contains the answer. Problem 1 is posed as a search over achievable stiffness values, and Section III describes physical apparatus plus a lookup table relating motor angle, stiffness, and compression; that lookup table is a hardware calibration, not a theoretical result that encodes the optimal stiffness. The phrase 'We argue that this steady-state hopping height is directly related to energy-efficient hopping' is a definitional choice of performance metric, not an inversion of the derivation: the optimum stiffness is not defined in terms of apex height but is discovered by experiment. The simulation is described as a validation and guidance tool rather than a first-principles prediction of the experimental ranking, and the supplied text ends before the results sections, so there is no exhibited evidence that simulation parameters were fitted to force the experimental trend. The author self-citations ([18], [25], [24], [36]) provide hardware components and the task state machine; they do not carry the load-bearing argument that stiffness should vary with ground profile. No equation-level circularity, fitted-input-as-prediction, or self-citation chain can be quoted from the paper. The skeptic's concern about whether the constant-energy protocol was actually enforced across stiffness settings (i.e., whether leg pre-compression was adjusted so that 0.5*k_l*delta^2 = E_in) is a potential experimental-validity issue, not a circularity, and the paper gives no quotation showing that the compared conditions were equalized by construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a small number of modeling and metric choices: a linear spring-damper ground, apex height as an energy-efficiency proxy, constant pre-compression energy across stiffness settings, and a simplified simulation. The free parameters are the discrete stiffness set, simulation-only leg damping values, and chosen input energies; none of these are derived from first principles.

free parameters (3)
  • Experimental leg stiffness selection = 3351, 4279, 5341 N/m
    The problem is solved only over this hand-picked set; the claimed best stiffness is optimal within these three values, not over a continuum.
  • Simulation leg damping coefficient = 30, 35, 40 Ns/m
    These damping values are selected for the simulation sweep. The real robot's intrinsic leg damping is not measured or reported in Table I, so simulation conclusions depend on these chosen values.
  • Input energy per hop = 0.97 J experiment; 1.0, 1.56, 2.25 J simulation
    The optimal stiffness map is conditioned on a fixed input energy; the paper does not test whether the ordering persists at other energy levels.
assumptions (4)
  • domain assumption The ground can be modeled as a parallel Hookean spring and a viscous damper with parameters (k_g, b_g).
    Used to define the ground profile in Section II and implemented by the ground emulator; real terrain may have nonlinear, plastic, or rate-dependent response not captured by this linear model.
  • ad hoc to paper Steady-state apex hopping height is directly related to energy-efficient hopping.
    Section II states this explicitly: 'We argue that this steady-state hopping height is directly related to energy-efficient hopping.' All optimality claims are in terms of apex height, not measured energy consumption.
  • domain assumption The pre-compression of the leg stores a constant input energy that is independent of the chosen leg stiffness.
    Section III.A describes a lookup table relating motor angle and leg stiffness to compression; if energy losses in the tendon or bellows depend on stiffness, the constant-energy condition may fail.
  • domain assumption The simplified mass-spring-damper simulation captures the dominant experimental dynamics.
    Section IV uses the same linear spring-damper structure; the paper asserts similar behavior with experiments, but the numerical comparison is not shown in the provided text.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tunable Leg Stiffness in a Monopedal Hopper for Energy-Efficient Vertical Hopping Across Varying Ground Profiles." pith.science (2026). https://pith.science/paper/EMTSB3N3

@misc{pith2026250802873,
  author       = {Pith},
  title        = {Pith review of: Tunable Leg Stiffness in a Monopedal Hopper for Energy-Efficient Vertical Hopping Across Varying Ground Profiles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EMTSB3N3}},
  note         = {Machine review of arXiv:2508.02873}
}
read the original abstract

We present the design and implementation of HASTA (Hopper with Adjustable Stiffness for Terrain Adaptation), a vertical hopping robot with real-time tunable leg stiffness, aimed at optimizing energy efficiency across various ground profiles (a pair of ground stiffness and damping conditions). By adjusting leg stiffness, we aim to maximize apex hopping height, a key metric for energy-efficient vertical hopping. We hypothesize that softer legs perform better on soft, damped ground by minimizing penetration and energy loss, while stiffer legs excel on hard, less damped ground by reducing limb deformation and energy dissipation. Through experimental tests and simulations, we find the best leg stiffness within our selection for each combination of ground stiffness and damping, enabling the robot to achieve maximum steady-state hopping height with a constant energy input. These results support our hypothesis that tunable stiffness improves energy-efficient locomotion in controlled experimental conditions. In addition, the simulation provides insights that could aid in the future development of controllers for selecting leg stiffness.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 34 canonical work pages

  1. [26]

    Learning to collab- orate with unknown agents in the absence of reward,

    Z. Zhang, H. Zhou, M. Imani, T. Lee, and T. Lan, “Learning to collab- orate with unknown agents in the absence of reward,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 13, 2025, pp. 14 502–14 511

    work page 2025

  2. [35]

    Modeling other players with bayesian beliefs for games with incomplete information,

    Z. Zhang, M. Imani, and T. Lan, “Modeling other players with bayesian beliefs for games with incomplete information,” arXiv preprint arXiv:2405.14122, 2024

    arXiv 2024

  3. [1]

    Interaction of leg stiffness and surface stiffness during human hopping,

    D. P. Ferris and C. T. Farley, “Interaction of leg stiffness and surface stiffness during human hopping,” Journal of applied physiology , vol. 82, no. 1, pp. 15–22, 1997

    work page 1997

  4. [2]

    Leg stiffness primarily depends on ankle stiffness during human hopping,

    C. T. Farley and D. C. Morgenroth, “Leg stiffness primarily depends on ankle stiffness during human hopping,” Journal of biomechanics , vol. 32, no. 3, pp. 267–273, 1999

    work page 1999

  5. [3]

    Running springs: speed and animal size,

    C. T. Farley, J. Glasheen, and T. A. McMahon, “Running springs: speed and animal size,” Journal of experimental Biology , vol. 185, no. 1, pp. 71–86, 1993

    work page 1993

  6. [4]

    Running in the real world: adjusting leg stiffness for different surfaces,

    D. P. Ferris, M. Louie, and C. T. Farley, “Running in the real world: adjusting leg stiffness for different surfaces,” Proceedings of the Royal Society of London. Series B: Biological Sciences , vol. 265, no. 1400, pp. 989–994, 1998

    work page 1998

  7. [5]

    Anymal-a highly mobile and dynamic quadrupedal robot,

    M. Hutter, C. Gehring, D. Jud, A. Lauber, C. D. Bellicoso, V . Tsounis, J. Hwangbo, K. Bodie, P. Fankhauser, M. Bloesch, et al., “Anymal-a highly mobile and dynamic quadrupedal robot,” in 2016 IEEE/RSJ international conference on intelligent robots and systems (IROS) . IEEE, 2016, pp. 38–44

    work page 2016

  8. [6]

    Design principles for a family of direct-drive legged robots,

    G. Kenneally, A. De, and D. E. Koditschek, “Design principles for a family of direct-drive legged robots,” IEEE Robotics and Automation Letters, vol. 1, no. 2, pp. 900–907, 2016

    work page 2016

Show all 36 references
  1. [7]

    Mit cheetah 3: Design and control of a robust, dynamic quadruped robot,

    G. Bledt, M. J. Powell, B. Katz, J. Di Carlo, P. M. Wensing, and S. Kim, “Mit cheetah 3: Design and control of a robust, dynamic quadruped robot,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 2245–2252

    2018

  2. [8]

    Actuator trans- parency and the energetic cost of proprioception,

    G. Kenneally, W.-H. Chen, and D. E. Koditschek, “Actuator trans- parency and the energetic cost of proprioception,” in Proceedings of the 2018 International Symposium on Experimental Robotics . Springer, 2020, pp. 485–495

    2018

  3. [9]

    An intro- ductory review of active compliant control,

    M. Schumacher, J. Wojtusch, P. Beckerle, and O. V on Stryk, “An intro- ductory review of active compliant control,” Robotics and Autonomous Systems, vol. 119, pp. 185–200, 2019

    2019

  4. [10]

    Design principles for energy-efficient legged locomotion and implementation on the mit cheetah robot,

    S. Seok, A. Wang, M. Y . Chuah, D. J. Hyun, J. Lee, D. M. Otten, J. H. Lang, and S. Kim, “Design principles for energy-efficient legged locomotion and implementation on the mit cheetah robot,” Ieee/asme transactions on mechatronics, vol. 20, no. 3, pp. 1117–1129, 2014

    2014

  5. [11]

    A review of quadruped robots: Structure, control, and autonomous motion,

    Y . Fan, Z. Pei, C. Wang, M. Li, Z. Tang, and Q. Liu, “A review of quadruped robots: Structure, control, and autonomous motion,” Advanced Intelligent Systems , p. 2300783, 2024

    2024

  6. [12]

    Next generation legged robot locomotion: A review on control techniques,

    S. S. Kotha, N. Akter, S. H. Abhi, S. K. Das, M. R. Islam, M. F. Ali, M. H. Ahamed, M. M. Islam, S. K. Sarker, M. F. R. Badal,et al., “Next generation legged robot locomotion: A review on control techniques,” Heliyon

  7. [13]

    Biorobotics: Using robots to emulate and investigate agile locomotion,

    A. J. Ijspeert, “Biorobotics: Using robots to emulate and investigate agile locomotion,” science, vol. 346, no. 6206, pp. 196–203, 2014

    2014

  8. [14]

    Dynamic walking: Toward agile and efficient bipedal robots,

    J. Reher and A. D. Ames, “Dynamic walking: Toward agile and efficient bipedal robots,” Annual Review of Control, Robotics, and Autonomous Systems, vol. 4, no. 1, pp. 535–572, 2021

    2021

  9. [15]

    Design and control of a tunable-stiffness coiled-spring actuator,

    S. Misra, M. Mitchell, R. Chen, and C. Sung, “Design and control of a tunable-stiffness coiled-spring actuator,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 634–640

    2023

  10. [16]

    Chapter 4 - control of motion and compliance,

    K. Mombaur, H. Vallery, Y . Hu, J. Buchli, P. Bhounsule, T. Boaventura, P. M. Wensing, S. Revzen, A. D. Ames, I. Poulakakis, and A. Ijspeert, “Chapter 4 - control of motion and compliance,” in Bioinspired Legged Locomotion, M. A. Sharbafi and A. Seyfarth, Eds. Butterworth- Hei...

    2017

  11. [17]

    Knee stiffness adjustment for energy efficient locomotion of a legged robot on surfaces with different stiffness,

    H. Q. Vu and L. G. Marcantini, “Knee stiffness adjustment for energy efficient locomotion of a legged robot on surfaces with different stiffness,” in 2013 IEEE International Conference on Robotics and Biomimetics (ROBIO). IEEE, 2013, pp. 1825–1831

    2013

  12. [18]

    Design and characteri- zation of a pneumatic tunable-stiffness bellows actuator,

    R. Chen, J. Kwon, W.-H. Chen, and C. Sung, “Design and characteri- zation of a pneumatic tunable-stiffness bellows actuator,” in2024 IEEE 7th International Conference on Soft Robotics (RoboSoft) . IEEE, 2024, pp. 997–1003

    2024

  13. [19]

    Adaptive control of quadruped locomotion through variable compliance of revolute spiral feet,

    A. Mutka, E. Ko ˇco, and Z. Kova ˇci´c, “Adaptive control of quadruped locomotion through variable compliance of revolute spiral feet,” Inter- national journal of advanced robotic systems , vol. 11, no. 10, p. 162, 2014

    2014

  14. [20]

    Compliance control of a legged robot based on improved adaptive control: Method and experiments,

    Y . Zhu and B. Jin, “Compliance control of a legged robot based on improved adaptive control: Method and experiments,” Int. J. Robot. Autom, vol. 5, pp. 366–373, 2016

    2016

  15. [21]

    Exper- imental investigations into the role of passive variable compliant legs for dynamic robotic locomotion,

    K. C. Galloway, J. E. Clark, M. Yim, and D. E. Koditschek, “Exper- imental investigations into the role of passive variable compliant legs for dynamic robotic locomotion,” in IEEE International Conference on Robotics and Automation , 2011, pp. 1243–1249

    2011

  16. [22]

    New variable passive-compliant element design for quadruped adaptation to stiffness-varying terrain,

    E. Koco, A. Mutka, and Z. Kovacic, “New variable passive-compliant element design for quadruped adaptation to stiffness-varying terrain,” International journal of advanced robotic systems, vol. 13, no. 3, p. 90, 2016

    2016

  17. [23]

    Uped: A quadruped robot platform to study directional leg compliance,

    J. Gurney, A. Samare Filsoofi, E. McClain, C. Bolejack, M. Abo- ufazeli, V . John Mathews, and S. Meek, “Uped: A quadruped robot platform to study directional leg compliance,” Journal of Mechanisms and Robotics, vol. 15, no. 1, p. 015001, 2023

    2023

  18. [24]

    Virtual energy management for physical energy savings in a legged robot hopping on granular media,

    S. F. Roberts and D. E. Koditschek, “Virtual energy management for physical energy savings in a legged robot hopping on granular media,” Frontiers in Robotics and AI , vol. 8, p. 740927, 2021

    2021

  19. [25]

    A tendon-driven origami hopper triggered by proprioceptive contact detection,

    W.-H. Chen, S. Misra, J. D. Caporale, D. E. Koditschek, S. Yang, and C. R. Sung, “A tendon-driven origami hopper triggered by proprioceptive contact detection,” in IEEE International Conference on Soft Robotics (RoboSoft) , 2020, pp. 373–380

    2020

  20. [27]

    A distributed abstract mac layer for cooperative learning on internet of vehicles,

    Y . Zou, Z. Zhang, C. Zhang, Y . Zheng, D. Yu, and J. Yu, “A distributed abstract mac layer for cooperative learning on internet of vehicles,” IEEE Transactions on Intelligent Transportation Systems , vol. 25, no. 8, pp. 8972–8983, 2024

    2024

  21. [28]

    Br-defedrl: Byzantine-robust decentralized federated reinforcement learning with fast convergence and communication efficiency,

    J. Qiao, Z. Zhang, S. Yue, Y . Yuan, Z. Cai, X. Zhang, J. Ren, and D. Yu, “Br-defedrl: Byzantine-robust decentralized federated reinforcement learning with fast convergence and communication efficiency,” in Ieee infocom 2024-ieee conference on computer com- munications. IEEE, ...

    2024

  22. [29]

    Distributed age-of-information scheduling with noma via deep reinforcement learning,

    C. Zhang, Y . Zou, Z. Zhang, D. Yu, J. T. G ´omez, T. Lan, F. Dressler, and X. Cheng, “Distributed age-of-information scheduling with noma via deep reinforcement learning,” IEEE Transactions on Mobile Com- puting, 2024

    2024

  23. [30]

    Byzantine fault tolerant consensus in open wireless networks via an abstract mac layer,

    G. Jing, Y . Zou, Z. Zhang, D. Yu, F. Dressler, and X. Cheng, “Byzantine fault tolerant consensus in open wireless networks via an abstract mac layer,” IEEE Transactions on Communications , 2024

    2024

  24. [31]

    Cooperative back- door attack in decentralized reinforcement learning with theoretical guarantee,

    M. Gao, Y . Zou, Z. Zhang, X. Cheng, and D. Yu, “Cooperative back- door attack in decentralized reinforcement learning with theoretical guarantee,” arXiv preprint arXiv:2405.15245 , 2024

    2024 arXiv

  25. [32]

    Adversarial inverse learning of defense policies conditioned on human factor models,

    A. Ravari, G. Jiang, Z. Zhang, M. Imani, R. H. Thomson, A. A. Pyke, N. D. Bastian, and T. Lan, “Adversarial inverse learning of defense policies conditioned on human factor models,” in 2024 58th Asilomar Conference on Signals, Systems, and Computers . IEEE, 2024, pp. 188–195

    2024

  26. [33]

    Network diffuser for placing- scheduling service function chains with inverse demonstration,

    Z. Zhang, V . Aggarwal, and T. Lan, “Network diffuser for placing- scheduling service function chains with inverse demonstration,” in IEEE INFOCOM 2025-IEEE Conference on Computer Communica- tions. IEEE, 2025, pp. 1–10

    2025

  27. [34]

    Look-ahead robust network optimization with generative state predictions,

    F. X. Yu, Z. Zhang, E. Grob, G. Adam, S. Coffey, N. D. Bastian, and T. Lan, “Look-ahead robust network optimization with generative state predictions,” in AAAI 2025 Workshop on Artificial Intelligence for Wireless Communications and Networking (AI4WCN)

    2025

  28. [36]

    Mitigating energy loss in a robot hopping on a physically emulated dissipative substrate,

    S. Roberts and D. E. Koditschek, “Mitigating energy loss in a robot hopping on a physically emulated dissipative substrate,” in 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 6763–6769

    2019

Pith tools

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