pith. sign in

arxiv: 2604.24006 · v1 · submitted 2026-04-27 · 📡 eess.SP

Efficient Near Field Beam Tracking via Thompson Sampling

Junchi Liu , Zijun Wang , Shawn Tsai , Rui Zhang This is my paper

Pith reviewed 2026-05-08 02:23 UTC · model grok-4.3

classification 📡 eess.SP
keywords beamrangetrackinganglebeamsdedicatedfieldframework
0
0 comments X

The pith

A Thompson sampling framework enables pilot-free joint angle-range beam tracking in the near field by modeling trajectories with low-order polynomials and using Fisher information for adaptive probing.

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

Large antenna arrays in future wireless networks create a near-field region where signals depend on both the direction and distance to the user. Traditional beam tracking interrupts data flow with special pilot signals sent in dedicated time slots. This work replaces those pilots with a decision strategy called Thompson sampling that picks beams likely to both track the user and carry useful data. It approximates the user's path over short time windows as simple curves in angle and distance, estimates the curve parameters from past observations, and measures uncertainty with a standard statistical tool called the Fisher information matrix. The sampling method then focuses future beams on the most uncertain parts of the path.

Core claim

Simulations demonstrate that the proposed framework maintains reliable connectivity while eliminating the overhead of dedicated pilot-based beam sweeping.

Load-bearing premise

User trajectories can be accurately captured by local low-order polynomials in angle and range over each sliding window, allowing maximum-likelihood estimation and Fisher-information uncertainty to guide Thompson sampling without dedicated pilots.

Figures

Figures reproduced from arXiv: 2604.24006 by Junchi Liu, Rui Zhang, Shawn Tsai, Zijun Wang.

Figure 2
Figure 2. Figure 2: Proposed beam tracking protocol. where L is the total number of propagation paths, the l-th scatter is located at polar coordinates (θl , rl,1) relative to the center of BS antenna array. From the geometry shown in view at source ↗
Figure 3
Figure 3. Figure 3: Tracking performance comparison of TS-based and pure exploitation view at source ↗
Figure 4
Figure 4. Figure 4: Tracking Performance under Different Feedback Symbol Ratio ( view at source ↗
read the original abstract

The shift to the radiative near field region due to large antenna arrays necessitates beamforming that accounts for both angle and range, evolving mobility management into a joint angular range tracking challenge. Conventional schemes rely on rigid pilot payload structures with dedicated training slots, which interrupt data transmission and degrade spectral efficiency. To address this, we propose a pilot-free beam tracking framework leveraging Thompson sampling(TS). Within each sliding window, the user trajectory is modeled by local low-order polynomials in angle and range, and the motion parameters are estimated by maximum likelihood with uncertainty quantified via the Fisher information matrix. TS adaptively probes uncertain trajectory regions using beams that simultaneously serve as payload beams. Simulations demonstrate that the proposed framework maintains reliable connectivity while eliminating the overhead of dedicated pilot-based beam sweeping.

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 pilot-free beam tracking framework for near-field massive MIMO systems. User trajectories are modeled as local low-order polynomials in angle and range over sliding windows; motion parameters are estimated via maximum-likelihood with uncertainty from the Fisher information matrix. Thompson sampling then adaptively selects beams that simultaneously carry payload data and probe uncertain trajectory regions. The central claim is that simulations show this approach maintains reliable connectivity while completely eliminating the overhead of dedicated pilot-based beam sweeping.

Significance. If the simulation results hold under realistic conditions, the work could meaningfully improve spectral efficiency in near-field regimes by removing pilot overhead. The approach correctly combines standard ML estimation, Fisher-information uncertainty quantification, and Thompson sampling without circularity in the adaptive choice. However, the absence of quantitative metrics, baselines, error bars, and robustness checks in the reported simulations limits the ability to assess practical impact.

major comments (2)
  1. [Simulations] Simulations section: the abstract and results provide no quantitative metrics (e.g., beam misalignment probability, achievable rate, or outage), no error bars, no baseline comparisons to pilot-based sweeping, and no details on how polynomial order, window length, or TS parameters were selected. This makes it impossible to evaluate the claimed 'reliable connectivity without overhead.'
  2. [Section 3] Trajectory modeling (Section 3): the central assumption that user motion is accurately captured by local low-order polynomials in angle and range over each sliding window is load-bearing for the pilot-free claim. No robustness experiments are reported against realistic deviations such as abrupt turns, higher-order dynamics, or non-smooth range changes; under model mismatch the ML estimates and Fisher uncertainty become biased, risking beam misalignment.
minor comments (2)
  1. [Section 3] Notation for the Fisher information matrix and Thompson sampling posterior sampling could be clarified with an explicit algorithmic pseudocode box.
  2. [Abstract and Simulations] The abstract states 'simulations demonstrate' but the results section should include at least one table or figure with concrete performance numbers and parameter settings.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and detailed assessment of our manuscript. We address each major comment below with clarifications and commit to revisions that strengthen the presentation of results and assumptions.

read point-by-point responses

  1. Referee: [Simulations] Simulations section: the abstract and results provide no quantitative metrics (e.g., beam misalignment probability, achievable rate, or outage), no error bars, no baseline comparisons to pilot-based sweeping, and no details on how polynomial order, window length, or TS parameters were selected. This makes it impossible to evaluate the claimed 'reliable connectivity without overhead.'

    Authors: We agree that additional quantitative detail is necessary to substantiate the performance claims. In the revised manuscript we will report explicit metrics including beam misalignment probability, achievable rate, and outage probability; include error bars on all simulation curves; add direct comparisons against conventional pilot-based sweeping baselines; and provide a dedicated subsection explaining the selection criteria and sensitivity analysis for polynomial order, sliding-window length, and Thompson sampling hyperparameters. These changes will enable a clearer evaluation of the overhead-free connectivity performance. revision: yes

  2. Referee: [Section 3] Trajectory modeling (Section 3): the central assumption that user motion is accurately captured by local low-order polynomials in angle and range over each sliding window is load-bearing for the pilot-free claim. No robustness experiments are reported against realistic deviations such as abrupt turns, higher-order dynamics, or non-smooth range changes; under model mismatch the ML estimates and Fisher uncertainty become biased, risking beam misalignment.

    Authors: The local low-order polynomial model over short sliding windows is a standard approximation in mobility tracking literature, motivated by the limited acceleration of typical user trajectories within the coherence time of near-field beams. Nevertheless, we acknowledge that explicit robustness checks are valuable. The revised version will incorporate new simulation results that evaluate tracking accuracy and beam misalignment under controlled model mismatches, including abrupt direction changes and higher-order motion components. This will quantify the degradation under mismatch and support the practical applicability of the pilot-free approach. revision: yes

Circularity Check

0 steps flagged

No circularity; standard statistical pipeline applied without self-referential reduction

full rationale

The derivation proceeds by assuming local low-order polynomial trajectories (an explicit modeling choice), applying maximum-likelihood estimation to recover parameters, using the Fisher information matrix to quantify uncertainty, and feeding that into Thompson sampling for beam selection. None of these steps defines the claimed performance metric (reliable connectivity without pilots) in terms of itself, nor renames a fitted quantity as a prediction, nor imports uniqueness via self-citation. Simulations test the method under its own modeling assumptions, which is standard validation rather than a circular reduction. The chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The framework rests on the domain assumption that short-term user motion is well-approximated by low-order polynomials and on standard statistical tools whose validity is taken from prior literature.

free parameters (2)
  • polynomial order
    Order of local polynomials for angle and range trajectories is chosen but not specified in abstract; affects model flexibility.
  • sliding window length
    Duration of each local modeling window is a design choice that trades off tracking responsiveness against estimation accuracy.
axioms (2)
  • domain assumption User trajectory can be locally approximated by low-order polynomials in angle and range.
    Invoked to enable maximum-likelihood parameter estimation within each sliding window.
  • standard math Fisher information matrix provides a reliable scalar measure of uncertainty for Thompson sampling decisions.
    Standard result from estimation theory, applied here to guide adaptive probing.

pith-pipeline@v0.9.0 · 5420 in / 1460 out tokens · 49530 ms · 2026-05-08T02:23:53.621998+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

  1. [1]

    A tutorial on near-field xl-mimo communications toward 6g,

    H. Lu et al., “A tutorial on near-field xl-mimo communications toward 6g,”IEEE Commun. Surveys Tuts., vol. 26, no. 4, pp. 2213–2257, Apr. 2024

    work page 2024

  2. [2]

    Communicating with extremely large-scale ar- ray/surface: Unified modeling and performance analysis,

    H. Lu and Y . Zeng, “Communicating with extremely large-scale ar- ray/surface: Unified modeling and performance analysis,”IEEE Trans. Wireless Commun., vol. 21, no. 6, pp. 4039–4053, Jun. 2022

    work page 2022

  3. [3]

    Channel estimation for extremely large-scale MIMO: Far-field or near-field?

    M. Cui and L. Dai, “Channel estimation for extremely large-scale MIMO: Far-field or near-field?”IEEE Transactions on Communica- tions, vol. 70, no. 4, pp. 2663–2677, 2022

    work page 2022

  4. [4]

    Tracking angles of departure and arrival in a mobile millimeter wave channel,

    C. Zhang, D. Guo, and P. Fan, “Tracking angles of departure and arrival in a mobile millimeter wave channel,” in2016 IEEE international conference on communications (ICC), IEEE, 2016, pp. 1–6

    work page 2016

  5. [5]

    Robust continuous-time beam tracking with liquid neural network,

    F. Zhu et al., “Robust continuous-time beam tracking with liquid neural network,” in2024 IEEE Global Communications Conference (GLOBECOM), IEEE, 2024

    work page 2024

  6. [6]

    Near-field beam tracking with extremely large dynamic metasurface antennas,

    P. Gavriilidis and G. C. Alexandropoulos, “Near-field beam tracking with extremely large dynamic metasurface antennas,”IEEE Transac- tions on Wireless Communications, 2025

    work page 2025

  7. [7]

    Beam training and tracking for extremely large-scale MIMO communications,

    K. Chen, C. Qi, C.-X. Wang, and G. Y . Li, “Beam training and tracking for extremely large-scale MIMO communications,”IEEE Transactions on Wireless Communications, vol. Early Access, 2024, Accepted for publication

    work page 2024

  8. [8]

    Fraunhofer and fresnel distances: Unified derivation for aperture antennas,

    K. T. Selvan and R. Janaswamy, “Fraunhofer and fresnel distances: Unified derivation for aperture antennas,”IEEE antennas and propa- gation magazine, vol. 59, no. 4, pp. 12–15, 2017

    work page 2017

  9. [9]

    Investigation of prediction accuracy, sensitivity, and parameter stability of large-scale propagation path loss models for 5g wireless communications,

    S. Sun et al., “Investigation of prediction accuracy, sensitivity, and parameter stability of large-scale propagation path loss models for 5g wireless communications,”IEEE transactions on vehicular technology, vol. 65, no. 5, pp. 2843–2860, 2016

    work page 2016

  10. [10]

    Next generation terahertz communications: A rendezvous of sensing, imag- ing, and localization,

    H. Sarieddeen, N. Saeed, T. Y . Al-Naffouri, and M.-S. Alouini, “Next generation terahertz communications: A rendezvous of sensing, imag- ing, and localization,”IEEE Commun. Mag., vol. 58, no. 5, pp. 69–75, May 2020

    work page 2020

  11. [11]

    Millimeter wave and sub-thz indoor radio propagation channel measurements, models, and comparisons in an office environment,

    Y . Xing, T. S. Rappaport, and A. Ghosh, “Millimeter wave and sub-thz indoor radio propagation channel measurements, models, and comparisons in an office environment,”IEEE Communications Letters, vol. 25, no. 10, pp. 3151–3155, 2021

    work page 2021

  12. [12]

    A tutorial on thompson sampling,

    D. Russo, B. Van Roy, A. Kazerouni, I. Osband, and Z. Wen, “A tutorial on thompson sampling,”F oundations and Trends in Machine Learning, vol. 11, no. 1, pp. 1–96, 2018

    work page 2018