pith. sign in

arxiv: 2607.00056 · v2 · pith:KTQYQUT2new · submitted 2026-06-30 · 💻 cs.IT · cs.AI· cs.MA· cs.NI· math.IT

Active Sensing for RIS-Aided Tracking and Power Control: A Hybrid Neuroevolution and Supervised Learning Approach

George Stamatelis , Hui Chen , Henk Henk Wymeersch , George C. Alexandropoulos This is my paper

Pith reviewed 2026-07-03 22:19 UTC · model grok-4.3

classification 💻 cs.IT cs.AIcs.MAcs.NImath.IT
keywords RISactive sensingtrackingpower controlneuroevolutiondual-agent learningsingle-bit feedback
0
0 comments X

The pith

A dual-agent neural framework jointly tunes RIS phases and user power for accurate energy-efficient tracking of mobile devices.

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

The paper introduces a Dual-Agent deep learning approach for active sensing that assists RIS-aided tracking of power-limited mobile users. It uses one agent to select discrete RIS phase profiles and another to decide uplink pilot power based on single-bit feedback from the base station. A hybrid training method combines neuroevolution with supervised learning to handle the non-differentiable nature of RIS elements and the limited feedback channel. Simulations show the method delivers robust tracking performance across multiple motion models and outperforms extended Kalman filters, particle filters, and other machine learning trackers, while also improving static localization over fingerprinting and reinforcement learning baselines.

Core claim

The Dual-Agent active sensing framework, trained by integrating the neuroevolution paradigm with supervised learning, jointly optimizes discrete RIS phase profiles and UE transmit power in real time. This overcomes non-differentiability of RIS responses and the strict bottleneck of single-bit feedback, enabling application to both single- and multi-antenna base stations and delivering highly accurate tracking that exceeds extended Kalman and particle filters as well as machine learning-based trackers.

What carries the argument

The Dual-Agent (DA) deep learning framework, where one neural network selects RIS phases and the second controls transmit power, trained via hybrid neuroevolution plus supervised learning.

If this is right

  • The same framework structure works for single-antenna and multi-antenna base stations with only an added output branch for combiner selection.
  • Single-bit feedback suffices for dynamic uplink power control while maintaining tracking accuracy.
  • The method reduces energy use by limiting pilot transmissions through real-time power adaptation.
  • Performance holds across diverse target motion models in numerical evaluations.

Where Pith is reading between the lines

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

  • The same hybrid training pattern could apply to other wireless problems that involve discrete hardware constraints and quantized feedback.
  • Extending the agents to multi-user scenarios would require only adding output dimensions for additional power decisions.
  • Hardware-in-the-loop tests with measured RIS phase responses would directly check whether the simulated gains survive real non-idealities.

Load-bearing premise

The hybrid neuroevolution and supervised learning training can successfully train the agents despite the non-differentiable discrete RIS phases and the single-bit feedback bottleneck.

What would settle it

A set of tracking simulations on the same motion models where the DA framework produces higher root-mean-square error than an extended Kalman filter or particle filter.

Figures

Figures reproduced from arXiv: 2607.00056 by George C. Alexandropoulos, George Stamatelis, Henk Henk Wymeersch, Hui Chen.

Figure 1
Figure 1. Figure 1: FIGURE 1: The considered RIS-aided active UE tracking [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 1
Figure 1. Figure 1: The considered RIS-aided active UE tracking system [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIGURE 2: Graphical illustration of the DA algorithm [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 2
Figure 2. Figure 2: Graphical illustration of the DA algorithm describing [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIGURE 3: A high-level overview of the estimator NN [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 3
Figure 3. Figure 3: A high-level overview of the estimator NN [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIGURE 4: The proposed policy NN for the multi [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 4
Figure 4. Figure 4: The proposed policy NN for the multi-antenna BS case. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIGURE 5: The considered example UE trajectories for different initial turn rates [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 5
Figure 5. Figure 5: The considered example UE trajectories for different initial turn rates [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIGURE 6: Trajectory RMSE performance over various UE motion and system parameters. [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 6
Figure 6. Figure 6: Trajectory RMSE performance over various UE motion and system parameters. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIGURE 7: Trajectory RMSE performance versus the [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
Figure 7
Figure 7. Figure 7: Trajectory RMSE performance versus the number of [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIGURE 8: Localization RMSE performance over various system parameters. [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 8
Figure 8. Figure 8: Localization RMSE performance over various system parameters. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIGURE 10: The average UE power levels in the uplink [PITH_FULL_IMAGE:figures/full_fig_p014_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Same as in Fig. 10, but for different observation [PITH_FULL_IMAGE:figures/full_fig_p012_11.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIGURE 14: Same as in Fig [PITH_FULL_IMAGE:figures/full_fig_p015_14.png] view at source ↗
Figure 10
Figure 10. Figure 10: The average UE power levels in the uplink of the [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIGURE 12: Sensitivity study results for tracking with [PITH_FULL_IMAGE:figures/full_fig_p015_12.png] view at source ↗
Figure 12
Figure 12. Figure 12: Sensitivity study results for tracking with [PITH_FULL_IMAGE:figures/full_fig_p013_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIGURE 13: Sensitivity study results for localization [PITH_FULL_IMAGE:figures/full_fig_p015_13.png] view at source ↗
Figure 13
Figure 13. Figure 13: Sensitivity study results for localization with noise [PITH_FULL_IMAGE:figures/full_fig_p013_13.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIGURE 15: Same as in Fig [PITH_FULL_IMAGE:figures/full_fig_p016_15.png] view at source ↗
read the original abstract

This paper studies energy efficient tracking of power-limited mobile users with the assistance of a Reconfigurable Intelligent Surface (RIS). Since localization pilot transmissions dominate the energy budget of power-constrained devices, we introduce a low-overhead feedback link from the Base Station (BS) to the user to enable dynamic uplink power control. To navigate the discrete and decentralized nature of this active sensing problem, we propose a novel Dual-Agent (DA) deep learning framework that jointly optimizes the discrete RIS phase profiles and the UE's transmit power in real time. Specifically, our approach employs a hybrid training methodology integrating the neuroevolution paradigm with supervised learning, effectively overcoming the non-differentiability of discrete phase responses from the RIS unit elements and the strict information bottleneck of single-bit feedback messages for pilot power control. The proposed DA active sensing framework can be applied with both single- and multi-antenna BSs, the latter with only minor modifications in the structure of one NN: an additional output branch with appropriate structure is included for the latter case to select a valid digital combiner from a finite set. Extensive numerical simulations demonstrate that the proposed scheme achieves highly accurate and robust tracking across diverse target motion models, outperforming extended Kalman and particle filters, as well as, machine learning-based trackers. Furthermore, in static localization, it is shown to significantly outperform traditional fingerprinting schemes, deep reinforcement learning baselines, and standard backpropagation-based estimators.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes a Dual-Agent (DA) deep learning framework for energy-efficient active sensing in RIS-aided tracking and uplink power control of power-limited mobile users. It jointly optimizes discrete RIS phase profiles and UE transmit power using a hybrid neuroevolution-supervised learning approach to handle non-differentiability and single-bit feedback constraints. The framework applies to single- and multi-antenna BSs (with minor NN modifications for the latter) and is evaluated via numerical simulations, claiming superior tracking accuracy and robustness over EKF, particle filters, and ML-based trackers across motion models, plus better static localization performance than fingerprinting, DRL, and backpropagation baselines.

Significance. If the simulation-based claims hold with adequate validation, the work would contribute to practical RIS-assisted localization by demonstrating a hybrid training method that navigates discrete optimization and limited feedback in real time. This could inform energy-efficient designs for power-constrained devices in future wireless systems, with the dual-agent structure and minor multi-antenna extension offering implementation flexibility. The absence of theoretical guarantees is offset by the emphasis on empirical performance across scenarios.

major comments (2)
  1. [Numerical Simulations] Numerical Simulations section: the abstract and results assert outperformance via extensive simulations, but the provided description supplies no details on simulation parameters (e.g., RIS size, SNR ranges, motion model specifics), datasets, exact metrics, or statistical significance testing; without these, it is impossible to assess whether the data support the central claims of robustness and superiority over EKF/particle filters and ML baselines.
  2. [Proposed DA Framework] Proposed DA Framework section: the hybrid neuroevolution + supervised learning methodology is presented as overcoming non-differentiability of discrete RIS phases and the single-bit feedback bottleneck, but the manuscript does not provide a concrete description or ablation showing how the neuroevolution component specifically resolves these issues (e.g., population size, mutation operators, or integration with the supervised loss); this is load-bearing for the performance claims.
minor comments (2)
  1. [Abstract] Abstract: the description of the multi-antenna extension is compressed; clarify the structure of the additional output branch for digital combiner selection.
  2. Notation: ensure consistent use of symbols for RIS phase profiles and power control variables across sections to avoid ambiguity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The two major comments identify areas where additional clarity will strengthen the manuscript. We address each point below and will incorporate the suggested expansions in the revised version.

read point-by-point responses

  1. Referee: [Numerical Simulations] Numerical Simulations section: the abstract and results assert outperformance via extensive simulations, but the provided description supplies no details on simulation parameters (e.g., RIS size, SNR ranges, motion model specifics), datasets, exact metrics, or statistical significance testing; without these, it is impossible to assess whether the data support the central claims of robustness and superiority over EKF/particle filters and ML baselines.

    Authors: We agree that the current description of the simulation setup is insufficient for independent verification. In the revised manuscript we will add a dedicated subsection (or expanded table) that explicitly lists all parameters: RIS dimensions, carrier frequency, SNR ranges, exact motion models and their parameters, channel models, performance metrics (RMSE, success rate, etc.), number of Monte Carlo runs, and any statistical significance tests performed. This will directly support the robustness and superiority claims. revision: yes

  2. Referee: [Proposed DA Framework] Proposed DA Framework section: the hybrid neuroevolution + supervised learning methodology is presented as overcoming non-differentiability of discrete RIS phases and the single-bit feedback bottleneck, but the manuscript does not provide a concrete description or ablation showing how the neuroevolution component specifically resolves these issues (e.g., population size, mutation operators, or integration with the supervised loss); this is load-bearing for the performance claims.

    Authors: We acknowledge that the manuscript would benefit from a more explicit algorithmic description of the neuroevolution component. In the revision we will expand the relevant subsection to include population size, mutation and crossover operators, selection mechanism, and the precise manner in which the neuroevolution output is combined with the supervised loss. We will also add a short ablation study isolating the contribution of the neuroevolution stage versus pure supervised learning, thereby clarifying how non-differentiability and the single-bit constraint are handled. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's central contribution is a Dual-Agent deep learning framework using hybrid neuroevolution and supervised learning to jointly optimize discrete RIS phases and UE transmit power. All performance claims (tracking accuracy, robustness across motion models, outperformance of EKF/particle filters/fingerprinting/DRL baselines) are presented as outcomes of extensive numerical simulations rather than closed-form derivations or predictions. No equations or steps reduce a claimed result to a fitted parameter or self-citation by construction; the hybrid training methodology is introduced as a practical solution to non-differentiability and single-bit feedback, with results framed as empirical evidence. The derivation chain is therefore self-contained against external simulation benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no identifiable free parameters, axioms, or invented entities; the central claims rest on the proposed hybrid training approach and simulation outcomes whose details are unavailable.

pith-pipeline@v0.9.1-grok · 5809 in / 1148 out tokens · 29681 ms · 2026-07-03T22:19:02.852601+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

74 extracted references · 3 canonical work pages · 3 internal anchors

  1. [1]

    Joint active RIS configuration and user power control for localization: A neuroevolution-based approach,

    G. Stamateliset al., “Joint active RIS configuration and user power control for localization: A neuroevolution-based approach,” inProc. IEEE ICASSP, (Barcelona, Spain), 2026

    2026

  2. [2]

    Localization methods for a mobile robot in urban environments,

    A. Georgiev and P. Allen, “Localization methods for a mobile robot in urban environments,”IEEE Trans. Robot., vol. 20, no. 5, pp. 851–864, 2004

    2004

  3. [3]

    SeqPolar: Sequence matching of polarized LiDAR map with HMM for intelligent vehicle localization,

    Q. Taoet al., “SeqPolar: Sequence matching of polarized LiDAR map with HMM for intelligent vehicle localization,”IEEE Trans. Veh. Technol., vol. 71, no. 7, pp. 7071–7083, 2022

    2022

  4. [4]

    A survey of the state-of-the-art localization techniques and their potentials for autonomous vehicle applications,

    S. Kuuttiet al., “A survey of the state-of-the-art localization techniques and their potentials for autonomous vehicle applications,”IEEE Intern. Things J., vol. 5, no. 2, pp. 829–846, 2018

    2018

  5. [5]

    Enabling optimization-based localization for IoT devices,

    P. N. Beuchatet al., “Enabling optimization-based localization for IoT devices,”IEEE Intern. Things J., vol. 6, no. 3, pp. 5639–5650, 2019

    2019

  6. [6]

    Survey of localization systems in internet of things,

    K. F. Bradaiet al., “Survey of localization systems in internet of things,” Mobile Netw Appl, vol. 24, p. 761–785, 2019

    2019

  7. [7]

    High-accuracy localization for assisted living: 5G systems will turn multipath channels from foe to friend,

    K. Witrisalet al., “High-accuracy localization for assisted living: 5G systems will turn multipath channels from foe to friend,”IEEE Signal Process. Mag., vol. 33, no. 2, pp. 59–70, 2016

    2016

  8. [8]

    Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,

    E. Basar, G. C. Alexandropoulos, Y . Liu, Q. Wu, S. Jin, C. Yuen, O. A. Dobre, and R. Schober, “Reconfigurable intelligent surfaces for 6G: Emerging hardware architectures, applications, and open challenges,” IEEE Veh. Technol. Mag., vol. 19, no. 3, pp. 27–47, 2024

    2024

  9. [9]

    Reconfigurable intelligent surface aided wireless local- ization,

    Y . Liuet al., “Reconfigurable intelligent surface aided wireless local- ization,” inProc. IEEE ICC, (Montreal, Canada), 2021

    2021

  10. [10]

    Near-field localization with a reconfigurable intelligent surface acting as lens,

    Z. Abu-Shabanet al., “Near-field localization with a reconfigurable intelligent surface acting as lens,” inProc. IEEE ICC, (Montreal, Canada), 2021

    2021

  11. [11]

    Towards ubiquitous positioning by leveraging reconfig- urable intelligent surface,

    H. Zhanget al., “Towards ubiquitous positioning by leveraging reconfig- urable intelligent surface,”IEEE Commun. Let, vol. 25, no. 1, pp. 284– 288, 2021

    2021

  12. [12]

    Wireless fingerprinting localization in smart environments using reconfigurable intelligent surfaces,

    C. L. Nguyenet al., “Wireless fingerprinting localization in smart environments using reconfigurable intelligent surfaces,”IEEE Access, vol. 9, pp. 135526–135541, 2021

    2021

  13. [13]

    Leveraging RIS-enabled smart signal propagation for solving infeasible localization problems: Scenarios, key research directions, and open challenges,

    K. Keykhosraviet al., “Leveraging RIS-enabled smart signal propagation for solving infeasible localization problems: Scenarios, key research directions, and open challenges,”IEEE Veh. Technol. Mag., vol. 18, no. 2, pp. 20–28, 2023

    2023

  14. [14]

    RISs and sidelink communications in smart cities: The key to seamless localization and sensing,

    H. Chenet al., “RISs and sidelink communications in smart cities: The key to seamless localization and sensing,”IEEE Commun. Mag., vol. 61, no. 8, pp. 140–146, 2023

    2023

  15. [15]

    Reconfigurable intelligent surfaces for energy effi- ciency in wireless communication,

    C. Huanget al., “Reconfigurable intelligent surfaces for energy effi- ciency in wireless communication,”IEEE Trans. Wireless Commun., vol. 18, no. 8, pp. 4157–4170, 2019

    2019

  16. [16]

    Reconfigurable intelligent surfaces for rich scattering wireless communications: Recent experiments, chal- lenges, and opportunities,

    G. C. Alexandropouloset al., “Reconfigurable intelligent surfaces for rich scattering wireless communications: Recent experiments, chal- lenges, and opportunities,”IEEE Commun. Mag., vol. 59, no. 6, pp. 28– 34, 2021

    2021

  17. [17]

    Towards smart and reconfigurable environment: Intelligent reflecting surface aided wireless network,

    Q. Wu and R. Zhang, “Towards smart and reconfigurable environment: Intelligent reflecting surface aided wireless network,”IEEE Commun. Mag., vol. 58, no. 1, pp. 106–112, 2020

    2020

  18. [18]

    RIS-enabled smart wireless environ- ments: Deployment scenarios, network architecture, bandwidth and area of influence,

    G. C. Alexandropouloset al., “RIS-enabled smart wireless environ- ments: Deployment scenarios, network architecture, bandwidth and area of influence,”EURASIP J. Wireless Commun. Netw., vol. 103, pp. 1–38, 2023

    2023

  19. [19]

    Asymptotically optimal closed-form phase configuration of 1-bit RISs via sign alignment,

    K. Stylianopouloset al., “Asymptotically optimal closed-form phase configuration of 1-bit RISs via sign alignment,” inProc. IEEE SPAWC, (Lucca, Italy), 2024

    2024

  20. [20]

    Reconfigurable intelligent surface assisted multiuser MISO systems exploiting deep reinforcement learning,

    C. Huanget al., “Reconfigurable intelligent surface assisted multiuser MISO systems exploiting deep reinforcement learning,”IEEE J. Sel. Areas Commun., vol. 38, no. 8, pp. 1839–1850, 2020

    2020

  21. [21]

    Online RIS configu- ration learning for arbitrary large numbers of 1-bit phase resolution elements,

    K. Stylianopoulos and G. C. Alexandropoulos, “Online RIS configu- ration learning for arbitrary large numbers of 1-bit phase resolution elements,” inProc. IEEE SPAWC, (Oulu, Finland), 2022

    2022

  22. [22]

    Pervasive machine learning for smart radio environments enabled by reconfigurable intelligent surfaces,

    G. C. Alexandropouloset al., “Pervasive machine learning for smart radio environments enabled by reconfigurable intelligent surfaces,”Proc. IEEE, vol. 110, no. 9, pp. 1494–1525, 2022

    2022

  23. [23]

    Deep contextual bandits for orchestrating multi-user MISO systems with multiple RISs,

    K. Stylianopouloset al., “Deep contextual bandits for orchestrating multi-user MISO systems with multiple RISs,” inProc. IEEE ICC, (Seoul, South Korea), 2022

    2022

  24. [24]

    Evolving multi-branch attention convolutional neural networks for online RIS configuration,

    G. Stamateliset al., “Evolving multi-branch attention convolutional neural networks for online RIS configuration,”IEEE Trans. Cogn. Commun. Netw., vol. 12, pp. 14–28, 2025

    2025

  25. [25]

    Sequential design of experiments,

    H. Chernoff, “Sequential design of experiments,”Ann. Math. Stat., vol. 30, no. 3, pp. 755–770, 1959

    1959

  26. [26]

    Chernoff test for strong-or-weak radar models,

    M. Franceschettiet al., “Chernoff test for strong-or-weak radar models,” IEEE Trans. Signal Process., vol. 65, no. 2, pp. 289–302, 2017

    2017

  27. [27]

    Active hypothesis testing for anomaly detec- tion,

    K. Cohen and Q. Zhao, “Active hypothesis testing for anomaly detec- tion,”IEEE Trans. Inf. Theory, vol. 61, no. 3, pp. 1432–1450, 2015

    2015

  28. [28]

    Deep active learning approach to adaptive beamform- ing for mmWave initial alignment,

    F. Sohrabiet al., “Deep active learning approach to adaptive beamform- ing for mmWave initial alignment,” inProc. IEEE ICASSP, (Toronto, Canada), 2021

    2021

  29. [29]

    Active hypothesis testing in un- known environments using recurrent neural networks and model free reinforcement learning,

    G. Stamatelis and N. Kalouptsidis, “Active hypothesis testing in un- known environments using recurrent neural networks and model free reinforcement learning,” inProc. EUSIPCO, (Helsinki, Finland), 2023

    2023

  30. [30]

    Active sensing for communications by learning,

    F. Sohrabiet al., “Active sensing for communications by learning,”IEEE J. Sel. Areas Commun, vol. 40, no. 6, pp. 1780–1794, 2022

    2022

  31. [31]

    Single- and multi-agent private active sensing: A deep neuroevolution approach,

    G. Stamateliset al., “Single- and multi-agent private active sensing: A deep neuroevolution approach,” inProc. IEEE ICC, (Denver, Colorado, USA), 2024

    2024

  32. [32]

    Localization with reconfigurable intelligent surface: An active sensing approach,

    Z. Zhanget al., “Localization with reconfigurable intelligent surface: An active sensing approach,”IEEE Trans. Wireless Commun., vol. 23, no. 7, pp. 7698–7711, 2024

    2024

  33. [33]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,”Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997

    1997

  34. [34]

    Lagrangian deep learning for private RIS-aided localization: An active sensing approach,

    G. Stamatelis and G. C. Alexandropoulos, “Lagrangian deep learning for private RIS-aided localization: An active sensing approach,” inProc. IEEE ICASSP, (Barcelona, Spain), 2026

    2026

  35. [35]

    Reconfigurable intelligent surfaces for wireless commu- nications: Overview of hardware designs, channel models, and estima- tion techniques,

    M. Jianet al., “Reconfigurable intelligent surfaces for wireless commu- nications: Overview of hardware designs, channel models, and estima- tion techniques,”Intell. Converg. Netw., vol. 3, no. 1, pp. 1–32, 2022

    2022

  36. [36]

    MIMO communications with 1-bit RIS: Asymp- totic analysis and over-the-air channel diagonalization,

    P. Gavriilidiset al., “MIMO communications with 1-bit RIS: Asymp- totic analysis and over-the-air channel diagonalization,” inProc. IEEE Asilomar, (Pacific Grove, USA), October 2025

    2025

  37. [37]

    Wideband sub-THz reconfigurable intelligent surface using planar tightly coupled dipoles,

    Y . Zhouet al., “Wideband sub-THz reconfigurable intelligent surface using planar tightly coupled dipoles,”IEEE Antennas and Wireless Propagation Letters, vol. 23, pp. 3659–3663, November 2024

    2024

  38. [38]

    Energy conservation in wireless sensor networks: A survey,

    G. Anastasiet al., “Energy conservation in wireless sensor networks: A survey,”Ad Hoc Netw., vol. 7, no. 3, pp. 537–568, 2009

    2009

  39. [39]

    Energy-aware wireless microsensor networks,

    V . Raghunathanet al., “Energy-aware wireless microsensor networks,” IEEE Signal Process. Mag., vol. 19, no. 2, pp. 40–50, 2002

    2002

  40. [40]

    Wireless integrated network sensors,

    G. J. Pottie and W. J. Kaiser, “Wireless integrated network sensors,” Commun. ACM, vol. 43, no. 5, p. 51–58, 2000

    2000

  41. [41]

    Bayesian user localization and tracking for reconfig- urable intelligent surface aided MIMO systems,

    B. Tenget al., “Bayesian user localization and tracking for reconfig- urable intelligent surface aided MIMO systems,”IEEE J. Sel. Top. Signal Process, vol. 16, no. 5, pp. 1040–1054, 2022

    2022

  42. [42]

    Performance limit and phase design for near-field tracking with reconfigurable intelligent surface,

    B. Wanget al., “Performance limit and phase design for near-field tracking with reconfigurable intelligent surface,”Digital Signal Process., vol. 167, p. 105426, 2025

    2025

  43. [43]

    Multi-user tracking in reconfigurable intelligent surface aided near-field wireless communications system,

    Y . Meiet al., “Multi-user tracking in reconfigurable intelligent surface aided near-field wireless communications system,”Appl. Sci., vol. 14, no. 1, 2024

    2024

  44. [44]

    Yu and M

    X. Yu and M. Gen,Introduction to Evolutionary Algorithms. Springer London, 2010

    2010

  45. [45]

    Efficient reinforcement learning through symbiotic evolution,

    D. E. Moriarty and R. Miikkulainen, “Efficient reinforcement learning through symbiotic evolution,”Mach. Learn., no. AI94-224, pp. 11–32, 1996

    1996

  46. [46]

    Co-evolving recurrent neurons learn deep memory POMDPs,

    F. J. Gomez and J. Schmidhuber, “Co-evolving recurrent neurons learn deep memory POMDPs,” inProc. GECCO, (Washington, DC, USA), 2005

    2005

  47. [47]

    Deep Neuroevolution: Genetic Algorithms Are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning

    F. Petroskiet al., “Deep neuroevolution: Genetic algorithms are a com- petitive alternative for training deep neural networks for reinforcement learning,”arXiv preprint: 1712.06567, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  48. [48]

    Evolution Strategies as a Scalable Alternative to Reinforcement Learning

    T. Salimanset al., “Evolution strategies as a scalable alternative to reinforcement learning,”arXiv preprint: 1703.03864, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  49. [49]

    Back to basics: Benchmarking canonical evolution strategies for playing atari,

    P. Chrabaszczet al., “Back to basics: Benchmarking canonical evolution strategies for playing atari,” inProc. IJCAI, (Stockholm, Sweden), 2018

    2018

  50. [50]

    Ultra-low power receivers for IoT applications: A review,

    D. D. Wentzloffet al., “Ultra-low power receivers for IoT applications: A review,” inProc. IEEE CICC, (Virtual), 2020

    2020

  51. [51]

    Ultra-low power wake-up radio for 5G IoT,

    A. Froytloget al., “Ultra-low power wake-up radio for 5G IoT,”IEEE Commun. Mag., vol. 57, no. 3, pp. 111–117, 2019

    2019

  52. [52]

    An approach to target tracking,

    M. Gruber, “An approach to target tracking,” technical report, MIT Lincoln Laboratory, Lexington, MA, 1967

    1967

  53. [53]

    Particle learning and smoothing,

    C. M. Carvalhoet al., “Particle learning and smoothing,”Statistical Science, vol. 25, no. 1, pp. 88–106, 2010

    2010

  54. [54]

    On the impact of control signaling in RIS- empowered wireless communications,

    F. Saggeseet al., “On the impact of control signaling in RIS- empowered wireless communications,”IEEE Open J. Commun. Soc, vol. 5, pp. 4383–4399, 2024

    2024

  55. [55]

    Analytic framework for the effective rate of MISO fading channels,

    M. Matthaiouet al., “Analytic framework for the effective rate of MISO fading channels,”IEEE Trans. Commun., vol. 60, no. 6, pp. 1741–1751, 2012. 15

    2012

  56. [56]

    Position aided beam alignment for millimeter wave backhaul systems with large phased arrays,

    G. C. Alexandropoulos, “Position aided beam alignment for millimeter wave backhaul systems with large phased arrays,” inProc. IEEE CAMSAP, (Curacao, Dutch Antilles), 2017

    2017

  57. [57]

    Noncoherent MIMO communications: Theoretical foundation, design approaches, and future challenges,

    K.-H. Ngoet al., “Noncoherent MIMO communications: Theoretical foundation, design approaches, and future challenges,”IEEE Open J. Veh. Technol., vol. 7, pp. 381–401, 2026

    2026

  58. [58]

    Near-field hierarchical beam management for RIS-enabled millimeter wave multi-antenna systems,

    G. C. Alexandropouloset al., “Near-field hierarchical beam management for RIS-enabled millimeter wave multi-antenna systems,” inProc. IEEE SAM, (Trondheim, Norway), 2022

    2022

  59. [59]

    Krishnamurthy,Partially Observed Markov Decision Processes: Fil- tering, Learning and Controlled Sensing

    V . Krishnamurthy,Partially Observed Markov Decision Processes: Fil- tering, Learning and Controlled Sensing. Cambridge University Press, 2nd ed., 2025

    2025

  60. [60]

    The complexity of Markov decision processes,

    C. H. Papadimitriou and J. N. Tsitsiklis, “The complexity of Markov decision processes,”Math. Operat. Research, vol. 12, no. 3, pp. 441– 450, 1987

    1987

  61. [61]

    The complexity of decentralized control of Markov decision processes,

    D. S. Bernsteinet al., “The complexity of decentralized control of Markov decision processes,”Math. Operat. Research, vol. 27, no. 4, pp. 819–840, 2002

    2002

  62. [62]

    NR; physical layer procedure for data (Release 15),

    3GPP, “NR; physical layer procedure for data (Release 15),” 2017

    2017

  63. [63]

    Accelerated neural evolution through cooperatively coevolved synapses,

    F. Gomezet al., “Accelerated neural evolution through cooperatively coevolved synapses,”J. Mach. Learn. Res., vol. 9, no. 31, pp. 937–965, 2008

    2008

  64. [64]

    Indoor path loss measurements and modeling in an open-space office at 2.4 ghz and 5.8 ghz in the presence of people,

    S. Kaddouriet al., “Indoor path loss measurements and modeling in an open-space office at 2.4 ghz and 5.8 ghz in the presence of people,” in Proc. IEEE PIMRC, (Bologna, Italy), 2018

    2018

  65. [65]

    Application of the extended Kalman filter to ballistic trajectory estimation,

    R. E. Larsonet al., “Application of the extended Kalman filter to ballistic trajectory estimation,”Stanford Research Institute, Tech. Rep., 1967

    1967

  66. [66]

    Sequential monte carlo methods under model uncer- tainty,

    I. Urtcagaet al., “Sequential monte carlo methods under model uncer- tainty,” inProc. IEEE SSP, (Palma de Mallorca, Spain), 2016

    2016

  67. [67]

    Asynchronous methods for deep reinforcement learning,

    V . Mnihet al., “Asynchronous methods for deep reinforcement learning,” inProc. ICML, (NY , USA), 2016

    2016

  68. [68]

    Benchmarking Batch Deep Reinforcement Learning Algorithms

    A. Rayet al., “Benchmarking safe exploration in deep reinforcement learning,”arXiv preprint: 1910.01708, 2019

    work page internal anchor Pith review Pith/arXiv arXiv 1910

  69. [69]

    Stable-baselines3: Reliable reinforcement learning implementations,

    A. Raffinet al., “Stable-baselines3: Reliable reinforcement learning implementations,”J. Mach. Learn. Research, vol. 22, no. 268, pp. 1–8, 2021

    2021

  70. [70]

    Pygad: An intuitive genetic algorithm python library,

    A. F. Gad, “Pygad: An intuitive genetic algorithm python library,” Multimedia Tools and Applications, pp. 1–14, 2023

    2023

  71. [71]

    Active sensing for localization with reconfigurable intelligent surface,

    Z. Zhanget al., “Active sensing for localization with reconfigurable intelligent surface,” inProc. IEEE ICC, (Rome, Italy), 2023

    2023

  72. [72]

    Multi-agent actor-critic with harmonic annealing pruning for dynamic spectrum access systems,

    G. Stamateliset al., “Multi-agent actor-critic with harmonic annealing pruning for dynamic spectrum access systems,” inProc. EUSIPCO, (Palermo, Italy), 2025

    2025

  73. [73]

    KalmanNet: Neural network aided kalman filtering for partially known dynamics,

    G. Revachet al., “KalmanNet: Neural network aided kalman filtering for partially known dynamics,”IEEE Trans. Signal Process., vol. 70, pp. 1532–1547, 2022

    2022

  74. [74]

    Filtering Markov jump systems with partially known dynamics: A model-based deep learning approach,

    G. Stamatelis and G. C. Alexandropoulos, “Filtering Markov jump systems with partially known dynamics: A model-based deep learning approach,”IEEE Trans. Signal Process., 2026 (to appear)

    2026