pith. sign in

arxiv: 2605.21815 · v2 · pith:5ASLIE34new · submitted 2026-05-20 · 💻 cs.IT · eess.SP· math.IT

Near-Field User Location Inference From Far-Field Power Measurements

Shima Mashhadi , Tiep M. Hoang , Alireza Vahid This is my paper

Pith reviewed 2026-05-22 07:31 UTC · model grok-4.3

classification 💻 cs.IT eess.SPmath.IT
keywords near-field beamfocusingpassive localizationpower measurementsbeam leakageuser location inferenceELAAnoncentral chi-square
0
0 comments X

The pith

Near-field beamfocusing leakage lets far-field power sensors infer user locations.

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

The paper shows that focusing energy at a chosen point near an extremely large array creates a structured leakage pattern whose strength at distant locations depends on the exact focal coordinates. Distributed sensors that record only received power can therefore extract a signature sufficient to recover the user's position. The authors derive a Bayesian lower bound on the resulting estimation error and demonstrate workable performance with both a grid-search estimator and a permutation-invariant neural network. This approach would matter in 6G networks because it converts an intended energy concentration into a passive localization resource that requires no phase information and no cooperation from the user.

Core claim

The radiated field outside the focal point exhibits a structured leakage that varies with the focal-point coordinates, enabling distributed far-field sensors to infer the user's location from received power measurements modeled by noncentral chi-square statistics.

What carries the argument

The location-dependent power signature produced by structured leakage outside the focal point, captured through noncentral chi-square statistics on far-field power measurements.

If this is right

  • Localization accuracy improves when more distributed sensors or additional measurement snapshots are available.
  • The method works under both line-of-sight and multipath channel conditions.
  • A classical grid-search estimator and an attention-based deep learning regressor both recover the location with usable precision.

Where Pith is reading between the lines

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

  • Network operators could monitor this leakage to track users passively, which raises questions about whether beamfocusing designs should deliberately suppress or shape the leakage.
  • The same signature might be combined with existing timing or angle measurements to tighten overall location estimates without extra hardware.
  • Extending the approach to moving users would require sequential updates of the power signature rather than single-shot inference.

Load-bearing premise

The leakage pattern must change distinctly enough with each possible user location that power readings at distant sensors can separate them reliably.

What would settle it

If the Bayesian Cramér-Rao lower bound shows persistently high error or if measured power distributions from different focal points overlap too much to support discrimination, the inference claim would not hold.

Figures

Figures reproduced from arXiv: 2605.21815 by Alireza Vahid, Shima Mashhadi, Tiep M. Hoang.

Figure 1
Figure 1. Figure 1: BCRLB versus number of passive FF sensors [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: NF beamfocusing leakage power map (dBm) over the 2D ar [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: evaluates the two estimators against the BCRLB for K ∈ {20, 40} sensors and snapshot lengths L = 1 and L = 50 across multiple noise powers. Both DeepSets and grid-search achieve finite localization error that improves systematically with more sensors (denser angular sampling of the leakage field) and more snapshots, while degrading with increasing noise as expected. DeepSets consistently outperforms the de… view at source ↗
Figure 4
Figure 4. Figure 4: Sionna-based performance under multipath propagat [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
read the original abstract

Near-field beamfocusing enabled by extremely large-aperture arrays (ELAA) is a promising 6G technique for massive connectivity and high spectrum efficiency. While beamfocusing concentrates energy at an intended user, the radiated field outside the focal point exhibits a structured leakage that varies with the focal-point coordinates. This paper shows that this leakage enables a new form of passive user localization in which distributed far-field sensors measuring only received power can infer the user's location by exploiting this location-dependent power signature. Using the induced noncentral chi-square statistics, we derive a Bayesian Cram\'er-Rao lower bound (BCRLB) that establishes the fundamental limits of this inference problem. We then evaluate a model-based grid-search estimator and an attention-based permutation-invariant deep learning regressor (DeepSet). Results under both line-of-sight (LoS) and multipath propagation confirm that reliable location inference is feasible, with accuracy improving as more sensors and snapshots are used.

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 claims that the structured leakage from near-field beamfocusing in extremely large-aperture arrays (ELAA) produces a location-dependent power signature that distributed far-field sensors can exploit for passive user localization using only received power measurements. It models the measurements with noncentral chi-square statistics, derives a Bayesian Cramér-Rao lower bound (BCRLB) to establish fundamental limits, and evaluates a model-based grid-search estimator alongside an attention-based permutation-invariant DeepSet regressor. Simulations under both line-of-sight and multipath conditions indicate that reliable inference is feasible, with accuracy improving as the number of sensors and snapshots increases.

Significance. If the central claim holds, the work introduces a new passive localization paradigm for 6G that leverages inherent properties of near-field beamfocusing without requiring active user transmission or specialized hardware. The BCRLB provides a theoretical benchmark, while the dual evaluation of grid-search and learning-based estimators demonstrates both analytical grounding and practical viability. Confirmation across LoS and multipath scenarios adds robustness, and the approach could inform future designs in massive connectivity and spectrum efficiency.

major comments (2)
  1. [BCRLB derivation and statistical model] The central feasibility claim rests on the expected power vector (under the noncentral chi-square model) being sufficiently unique and distinguishable across distinct focal-point locations. The manuscript does not include an explicit injectivity analysis, condition-number evaluation, or numerical sensitivity study of this mapping, particularly under multipath. Without such analysis (e.g., in the section deriving the BCRLB or the statistical model), the BCRLB may be optimistic and both estimators could encounter irreducible ambiguities in symmetric geometries or realistic channels.
  2. [Estimator evaluation and results] The simulation results confirm feasibility but do not report how closely the grid-search and DeepSet estimators approach the derived BCRLB as a function of sensor count or snapshots. Adding this comparison (e.g., in the results section or associated figures) would clarify whether performance is near-fundamental limits or limited by estimator sub-optimality.
minor comments (2)
  1. [Statistical model] Clarify the exact definition of the noncentrality parameter in the chi-square model and its dependence on the focal-point coordinates; this would improve reproducibility of the BCRLB.
  2. [Figures and results] Include the BCRLB curves as reference lines in the performance plots to facilitate direct visual comparison with estimator RMSE or error metrics.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment below and indicate the changes we will make in the revised version.

read point-by-point responses

  1. Referee: [BCRLB derivation and statistical model] The central feasibility claim rests on the expected power vector (under the noncentral chi-square model) being sufficiently unique and distinguishable across distinct focal-point locations. The manuscript does not include an explicit injectivity analysis, condition-number evaluation, or numerical sensitivity study of this mapping, particularly under multipath. Without such analysis (e.g., in the section deriving the BCRLB or the statistical model), the BCRLB may be optimistic and both estimators could encounter irreducible ambiguities in symmetric geometries or realistic channels.

    Authors: We agree that an explicit numerical sensitivity analysis would strengthen the support for the central claim. The BCRLB derivation relies on the statistical model being identifiable for the chosen parameterization, and the reported simulations under both LoS and multipath already show consistent estimator behavior across tested locations. To directly address the concern, we will add a new subsection (or appendix) containing a condition-number evaluation of the mapping from focal-point coordinates to the expected power vector, together with a Monte Carlo sensitivity study under multipath to quantify potential ambiguities in symmetric geometries. This material will be placed near the BCRLB derivation. revision: yes

  2. Referee: [Estimator evaluation and results] The simulation results confirm feasibility but do not report how closely the grid-search and DeepSet estimators approach the derived BCRLB as a function of sensor count or snapshots. Adding this comparison (e.g., in the results section or associated figures) would clarify whether performance is near-fundamental limits or limited by estimator sub-optimality.

    Authors: We thank the referee for this suggestion. While the BCRLB is presented as a benchmark, the current figures do not directly overlay estimator MSE against the bound. We will revise the results section to include additional plots (or a table) that compare the mean-squared error of both the grid-search and DeepSet estimators to the BCRLB, parameterized by the number of sensors and the number of snapshots. This will make explicit how close the practical estimators operate to the theoretical limit. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper derives the BCRLB directly from the noncentral chi-square distribution of the power measurements under the stated propagation model, then separately evaluates grid-search and DeepSet estimators on simulated data. No step reduces a claimed prediction or bound to a fitted parameter or self-citation by construction; the statistical model and performance metrics remain independently grounded in the measurement statistics and array geometry.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests primarily on the domain assumption of structured, location-dependent leakage in the radiated field and the applicability of noncentral chi-square statistics to power measurements. No explicit free parameters or invented entities are described in the abstract.

axioms (2)
  • domain assumption The radiated field outside the focal point exhibits a structured leakage that varies with the focal-point coordinates.
    This premise enables the location-dependent power signature used for inference.
  • domain assumption Power measurements follow induced noncentral chi-square statistics.
    This statistical model underpins the derivation of the Bayesian CRLB.

pith-pipeline@v0.9.0 · 5704 in / 1496 out tokens · 54996 ms · 2026-05-22T07:31:54.395880+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.