Near-Field Angular-Error Amplification

• Near-field angular-error amplification is a phenomenon where minor Cartesian uncertainties in the near field result in significantly amplified angular errors following a 1/r scaling law.
• The effect challenges robust beamforming and channel estimation in reconfigurable antenna arrays, potentially causing severe performance degradation or, alternatively, enabling ultra-sensitive detection.
• Specialized LMI-based designs and adaptive decoding strategies are employed to counteract these errors, improving both RF system reliability and precision in optical metrology.

The near-field angular-error amplification effect is a phenomenon in electromagnetic wave systems, beamforming, and precision measurement whereby small Cartesian domain errors or physical perturbations in the near field induce disproportionately large angular errors. This amplification is particularly prominent in scenarios involving spherical-wavefront propagation, location uncertainty, or field-pattern perturbations at sub-wavelength to Fresnel-scale distances. In the context of near-field beamforming and physically reconfigurable antenna arrays, as well as in weak-value optical metrology, such amplification can severely degrade system performance or alternately enable ultra-sensitive detection, depending on whether the effect is seen as a challenge or resource.

1. Mathematical Origin: Near-Field Channel Model and Error Mapping

In array systems operating in the electromagnetic near field, the channel from an antenna array to a target at position $(x, y)$ is most accurately described using a spherical-wave (Fresnel-regime) steering vector. For a uniform linear array (ULA) aligned along the $y$-axis and element positions $u_n = (0, u_n)$, the normalized steering vector toward position $(r, \theta)$ is

$$\mathbf{a}(r, \theta) = \left[ a_1(r, \theta), \cdots, a_N(r, \theta) \right]^T,$$

with

$$a_n(r, \theta) = \frac{1}{\sqrt{N}} \exp \left\{ j \frac{2\pi}{\lambda} \left[ u_n \sin\theta - \frac{u_n^2 \cos^2\theta}{2r} \right] \right\}.$$

This derives from the Fresnel-expanded distance

$$\left\| (r\cos\theta, r\sin\theta) - u_n \right\| - r \approx -u_n \sin\theta + \frac{u_n^2 \cos^2\theta}{2r}.$$

When the true location $(x, y)$ differs from a nominal estimate $(\hat{x}, \hat{y})$ by errors $(\Delta x, \Delta y)$, expansion to first order in $(\Delta x, \Delta y)$ yields

$$\Delta r = \frac{\hat{x} \Delta x + \hat{y} \Delta y}{r}, \qquad \Delta\theta \approx \frac{ \hat{x} \Delta y - \hat{y} \Delta x }{r^2 }.$$

Thus, a given Euclidean (Cartesian) error $\|\Delta \mathbf{q}\|$ induces a polar angle error $\Delta\theta$ scaling as $1/r$:

$$|\Delta\theta| \leq \frac{ \|\Delta \mathbf{q}\| }{ r }.$$

As $r \to 0$, the angular error becomes arbitrarily large for fixed Cartesian uncertainty (Zhou et al., 20 Jan 2026).

2. Phenomenological Manifestation: Amplification Across Physical Systems

The near-field angular-error amplification effect emerges in multiple domains:

• Robust beamforming under location uncertainty: In large-scale array systems, when eavesdropper (Eve) locations are uncertain, a conventional Cartesian uncertainty ball maps into a dominant angular error for nearby targets, drastically increasing worst-case steering-vector mismatch (Zhou et al., 20 Jan 2026).
• Beam-space MIMO perturbation: In beam-space MIMO with reconfigurable antennas encoding data in radiation patterns, near-field objects (e.g., a user's hand) perturb pattern phases differently for each constellation state. This yields symbol-specific angular deviation in the reconstructed basis, so that minor near-field reactance translates—after far-field demapping—into tens of degrees of virtual symbol rotation (Yousefbeiki et al., 2016).

The effect is universal to signal-processing frameworks wherein polar or angular encoding interacts with position or pattern uncertainties.

3. Impact on System Performance and Modeling

In wireless security and beamforming, the effect causes significant reduction in achievable sum-rate under robust null-steering when conventional methods are applied. Specifically:

• For a fixed uncertainty $\|\Delta\mathbf{q}\|$, the worst-case steering-vector perturbation $\varepsilon$ grows rapidly as the nominal range $\hat{r}$ decreases (e.g., $\varepsilon \approx 0.05$ at $\hat{r} = 15$ m and $\varepsilon > 0.3$ at $\hat{r} = 5$ m for $\sigma_c = 0.1$ m) (Zhou et al., 20 Jan 2026).
• Conventional error-bound LMIs must account for this enlarged uncertainty region, leading to highly conservative power allocations and annihilating beamforming gain at short ranges (Zhou et al., 20 Jan 2026).
• In pattern-reconfigurable MIMO, minor near-field perturbations ($|\Psi-1| = 0.1\ldots0.2$ for separations of a few millimeters) induce local symbol error vector magnitudes (EVMs) as bad as $-8$ dB and constellation rotations exceeding 30° for QPSK, with zero-forcing MIMO decoding catastrophically amplifying the mismatch (Yousefbeiki et al., 2016).

A plausible implication is that robust signal-processing solutions must directly model and mitigate angular-error amplification, rather than naively bounding overall steering-vector uncertainty.

4. Methodological Solutions: Specialized LMI and Decoding Architectures

To address this effect, recent work proposes two central strategies:

• Two-stage LMI-based beamforming: Partition the angular–radial uncertainty space into fan-shaped subregions (“sectors”) ensuring that locally $|\Delta\theta| \leq 1/(2N)$ and $\Delta r/r \ll 1$ hold. Within each sector, use a first-order Taylor expansion of the steering vector and enforce refined $4\times4$ LMI constraints derived from a Generalized S-Procedure, instead of a single $(N+2)\times(N+2)$ LMI on the full perturbation region (Zhou et al., 20 Jan 2026).
• EVM-driven adaptive decoding (MIMO): Continuously monitor EVM at the receiver; if EVM exceeds a chosen threshold (e.g., $-15$ dB), estimate and compensate angular shift $\delta\phi$ through pilot sequences and apply a corrective rotation before symbol demodulation. Alternatively, adaptively re-tune antenna reactances, using low-rate feedback, to equalize perturbation phases across pattern states, targeting $|\Psi^{\{+j\}}-\Psi^{\{-j\}}|<0.05$ for reliable QPSK operation (Yousefbeiki et al., 2016).

These methods explicitly exploit knowledge of the $1/r$ amplification law and first-order steering-vector sensitivity to restore robust system performance.

5. Analogous Optical Phenomena: Amplification via Weak Measurements

Structural analogs of angular-error amplification are observed in optical weak-value amplification protocols, both for angular rotation sensing and Goos–Hänchen beam shifts:

• Azimuthal weak-value amplification: Spin–orbit coupling in Sagnac interferometers produces a weakly entangled polarization–angle degree of freedom. By near-orthogonal post-selection of polarization, one achieves an amplification factor $M \approx 1/\gamma$ (where $\gamma \ll 1$ is a small analyzer angle) of the angular pointer shift, with observed gains up to $M \sim 100$ (Magana-Loaiza et al., 2013).
• Goos–Hänchen angular shift: At air–dielectric interfaces, tiny polarization-dependent angular deviations on reflection are amplified by selecting analyzer angles $\varepsilon \ll 1$, with the observed shift scaling as $1/\varepsilon$. This enables detection of micro-radian-scale beam deflections (Jayaswal et al., 2014).

These metaphors demonstrate that near-field angular-error amplification is not only detrimental (as in array beamforming), but can be engineered to generate sizable, detectable pointer shifts in precision metrology.

6. Experimental Observations and Practical Consequences

In robust near-field beamforming, measured error bounds verify that steering-vector uncertainty $\varepsilon$ escalates rapidly as target distance $\hat{r}$ decreases for constant Cartesian position uncertainty. For $\sigma_c = 0.1$ m, $\varepsilon$ increases from $0.05$ at $15$ m to $0.3$ at $5$ m, inducing angular errors from $0.0033$ rad ($0.19$°) to $0.06$ rad ($3.4$°), representing a near $20\times$ amplification (Zhou et al., 20 Jan 2026). Conventional robust design collapses under this uncertainty, but refined sectorized LMI approaches recover performance with limited complexity.

In beam-space MIMO, numerical results for a typical setup ($2.45$ GHz QPSK, single element, reconfigurable loads) show that local symbol-dependent EVM can spike above $-8$ dB with $\delta\phi$ exceeding $30$° for hand-antenna separations below $10$ mm. Adaptive basis recalibration or reactive-load optimization is necessary to maintain EVM $< -20$ dB (Yousefbeiki et al., 2016).

Optical weak-value schemes exploiting this effect achieve enhanced measurement sensitivities down to $10^{-6}$ rad of angular deviation, with amplification factors tunable by the post-selection angle $\gamma$ or $\varepsilon$ (Magana-Loaiza et al., 2013, Jayaswal et al., 2014).

7. Applications, Limitations, and Theoretical Implications

Near-field angular-error amplification is intrinsic to any application in which angular resolution or pattern discrimination under spatial uncertainty is critical. It imposes strong limitations on classical robust beamforming, channel estimation, and angular-multiplexed communications in the near field. Conversely, it enables ultra-sensitive angular displacement detection in quantum-inspired and classical optical metrology.

While the effect is a direct consequence of coordinate transformation—the $1/r$ scaling from Cartesian to polar errors—its impact is magnified in scenarios with high pattern directivity, dense arrays, or strong field-pattern–object coupling.

A plausible implication is that future system designs in both RF and photonic domains must incorporate explicit angular-error modeling at the design stage, adopting non-Euclidean uncertainty sets or leveraging amplification phenomena for precision tasks. Failure to do so results either in excessively conservative performance or unmitigated vulnerability to short-range adversaries.

---

Key References:

• "Near-field Physical Layer Security: Robust Beamforming under Location Uncertainty" (Zhou et al., 20 Jan 2026)
• "Amplification of Angular Rotations Using Weak Measurements" (Magana-Loaiza et al., 2013)
• "Near-Field Perturbation Effect on Constellation Error in Beam-Space MIMO" (Yousefbeiki et al., 2016)
• "Observing angular deviations in light beam reflection via weak measurements" (Jayaswal et al., 2014)