Most advice stops at “keep some distance—power drops with the inverse-square law.” That’s true once you’re in the far field. But phones (and your body) live a lot of the time in the Fresnel (radiating near-field) region, where fields are structured and distance-dependent, not purely 1/r². Understanding that middle zone explains why centimeters matter and why 5–15 mm test spacings aren’t a biological boundary.

1) The three regions—where 1/r² applies (and where it doesn’t)

For a radiator with maximum dimension D at wavelength λ, the space around it divides into:

• Reactive near-field: stored (non-radiating) energy dominates; E and H are not in phase.

• Fresnel (radiating near-field): radiation dominates but pattern and amplitude still change with distance due to phase curvature and aperture interference.

• Far field (Fraunhofer): fields locally plane-wave; E and H in phase and power ∝ 1/r². Formal boundaries often used:
  rRNF≈0.62D3/λr_{\text{RNF}} \approx 0.62 \sqrt{D^{3}/\lambda}rRNF​≈0.62D3/λ​ and rFF≈2D2/λr_{\text{FF}} \approx 2 D^{2}/\lambdarFF​≈2D2/λ. LSO+2Jontalle+2

Implication: 1/r² is reliably valid beyond rFFr_{\text{FF}}rFF​. Between rRNFr_{\text{RNF}}rRNF​ and rFFr_{\text{FF}}rFF​, interference and phase curvature create hot-spots/nulls and coupling that doesn’t follow a clean inverse square. Purdue Engineering

2) How “big” is the Fresnel region for a phone?

Because a handset’s effective aperture is often the chassis (≈ 8–12 cm) at sub-6 GHz, the Fresnel region typically runs centimeters to tens of centimeters—not millimeters. Example free-space estimates (D≈10 cm):

• 900 MHz: Fresnel ≈ ~2–6 cm

• 2.4 GHz: ~6–16 cm

• 3.5 GHz: ~7–23 cm
  At mmWave, modules are much smaller (e.g., D≈2 cm), so despite the shorter λ the region can be a few centimeters. Body proximity and arrays can extend these spans. (See the live table I generated for publish-ready numbers.) Jontalle

Why it grows with frequency (for the same D): rFF∝D2/λr_{\text{FF}}\propto D^{2}/\lambdarFF​∝D2/λ. Smaller λ ⇒ larger rFFr_{\text{FF}}rFF​. If D shrinks (mmWave modules), that offsets the effect.

3) Distance still helps—a lot (but don’t oversell 1/r² in the near field)

Once you’re past rFFr_{\text{FF}}rFF​, doubling distance ⇒ ¼ the power density (1/r²). Inside the Fresnel region, distance still reduces coupling, but not as neatly as 1/r² because E/H phase and aperture interference matter. That’s exactly the zone where interferometric shaping (like your Interferometric Ray Antenna concept) can focus or redirect energy—which cuts both ways for exposure and performance. Purdue Engineering

See the attached chart: it marks rRNFr_{\text{RNF}}rRNF​ and rFFr_{\text{FF}}rFF​ for a 2.4 GHz, D=10 cm example and shows where 1/r² applies cleanly.

4) Why “far field is harmless” is not supported by the animal evidence

Two large independent programs found tumor signals under whole-body (far-field–like) exposures:

• NTP (USA): “Clear evidence” of malignant heart schwannomas in male rats; some evidence for brain gliomas. National Toxicology Program+1

• Ramazzini Institute (Italy): Life-span exposures to 1.8 GHz GSM base-station-like fields (19 h/day) observed increased heart schwannomas in male rats—directionally consistent with NTP. PubMed+2ScienceDirect+2

Bottom line: Even without the structured coupling of the Fresnel zone, far-field RFR produced biological signals in animals under controlled conditions.

5) Signal structure matters: pulsing/modulation & plausible mechanisms

Real-life wireless always rides low-frequency envelopes on RF carriers (e.g., GSM ~217 Hz, DECT ~100–200 Hz, many 3G/4G/5G frames ~100 Hz). Multiple reviews argue that modulated/pulsed fields are more bioactive than unmodulated ones and outline mechanisms involving voltage-gated ion channels (VGIC/VGCC) and oxidative stress (ion forced-oscillation / reciprocating Lorentz-force concepts). These are active scientific debates, not settled consensus, but they are mechanistically plausible and consistent with many “effect” studies. PMC+1

6) Why the “5–15 mm” in manuals isn’t a safety distance

That small spacing is a test configuration used to show a phone can meet SAR limits under standardized conditions (IEC/IEEE 62209-1528). Many manufacturers historically specify ~5–15 mm for body-worn tests; some Samsung pages still list 10–15 mm or 1 cm separations for compliance. These are not biology boundaries—just as-tested spacings. More distance still reduces coupling. Samsung+1

7) Practical guidance (device-agnostic)

• Use distance strategically. For sub-6 GHz, ~10–30 cm typically moves you out toward/into the far field; for small mmWave modules, ~5–10 cm often suffices in free space (body loading can extend this). Jontalle

• Cut duty cycle. Batch downloads/streams; message, don’t stream audio next to the body.

• Prefer speakerphone/air-tube. Keep metal plates/magnets away from antenna areas (they can detune and force higher transmit power).

• Indoors: Prefer wired/Li-Fi when you can.

• Remember: Regulatory guidelines (e.g., ICNIRP 2020) are mostly thermal; they do not resolve all reported non-thermal endpoints. ICNIRP+1