RF exposure metrics above 6 GHz

Many RF safety misunderstandings above ~6 GHz are not about “stronger vs weaker RF” — they’re about metric definition. The quantity you compare to a framework often becomes an area-averaged surface heating metric, and point-probe intuition breaks.

Next action: If this is a real site assessment, pair this with Q. Uncertainty.

---

1) What changes around the ~6 GHz transition (physics, not hype)

As frequency increases, penetration depth generally decreases and absorption becomes more surface-localized. For close-to-body exposure, many frameworks therefore shift emphasis:

• from mass-averaged SAR (volumetric absorption) to
• area-averaged surface power deposition metrics (superficial heating control).

This does not mean “SAR stops existing”; it means the framework’s primary compliance metric can change because the heating geometry changes.

---

2) The three quantities people mix up

Terminology varies by framework, but the conceptual buckets are stable:

A) Incident power density (external; in air)

Power flow in air toward a surface. In a plane wave:

This conversion is a plane-wave assumption and can fail in the near field.

B) Absorbed power density (surface/tissue; internal dose concept)

How much power per unit area is actually deposited in tissue near the surface. This is not generally measured by a single “room meter” reading; it’s defined by the framework and often supported by standardized measurement/modelling methods.

C) Epithelial / surface power density (area-averaged; framework-defined)

Many frameworks define a specific averaging area (cm²-scale) for the comparison quantity. The exact area definition (and any additional smaller-area checks for focusing) is framework- and band-dependent.

Discipline rule: do not invent your own averaging area. Use what your adopted framework defines.

---

3) Why a point peak can be misleading (and what to do instead)

Above ~6 GHz (and for many beamformed systems), spatial structure can be sharp:

• small position shifts can change the reading,
• reflections create local maxima/minima,
• beams move the maximum in time and space,
• a probe’s physical size can be comparable to spatial features.

If your comparison metric is area-averaged, a defensible workflow is:

1. locate candidate maxima (scan)
2. define the framework-required averaging region around the maxima
3. compute the area-averaged value using a documented sampling plan
4. document uncertainty and sensitivity (positioning and time variability)

---

4) Measurement method notes (what to state in your report)

When working above ~6 GHz, you should explicitly state:

• whether the point is plausibly near field (arrays can push far field out to meters)
• whether you are comparing to incident or absorbed/surface criteria
• the averaging definition (time and area) and how your instrument/method implements it
• the operating state assumptions (beam states, duty, traffic, worst-case bounding)

If you can’t justify these, your result is screening only.

---

5) Common mistakes (and the corrected framing)

1. Using S = E²/377 close to an array
   Correct: classify the region and avoid plane-wave conversions in near field unless your method standard supports it.

2. Treating a needle peak as “the metric” when the metric is area-averaged
   Correct: compute the framework-defined area average after locating the maximum region.

3. Equating incident and absorbed criteria
   Correct: use the framework-defined relationship or perform refined evaluation; don’t swap them casually.

4. Max-hold screenshots used as time averages
   Correct: apply the framework-defined time window and (where relevant) worst-case window placement.

---

6) Further reading (primary sources you must edition-control)

• ICNIRP (2020): Guidelines for limiting exposure to electromagnetic fields (100 kHz to 300 GHz) — use the exact edition adopted by your jurisdiction, including definitions for basic restrictions vs reference levels and the high-frequency metric definitions.
• IEEE Std C95.1 (edition varies by regulator/contract) — use the exact revision required; apply DRL/ERL logic and the high-frequency metric definitions as specified.
• IEC 62232 (site assessment methods near radiocommunication installations) — method guidance for measurement and modeling, especially where near-field and multi-source complexity matter.

---

Knowledge check

Check your understanding 1

Why do many frameworks shift away from SAR as the primary metric above ~6 GHz?

Show answer

Because absorption becomes more superficial and surface heating becomes the dominant concern; frameworks often define surface/area-averaged metrics that better map to that geometry.

Check your understanding 2

Why does an incident power-density reading in air not directly give absorbed power density in tissue?

Show answer

Incident power density describes incoming power per unit area; absorbed power density describes deposition in tissue. Reflection, tissue properties and exposure geometry affect how much is absorbed. Use the adopted framework’s relationship or a valid dosimetric evaluation rather than relabelling the meter reading.

Check your understanding 3

Your criterion uses an area average. After locating a candidate hotspot, how should you measure the comparison quantity?

Show answer

Sample across the framework-defined averaging region using a documented spatial plan, then calculate that area’s average. State how the time averaging and source operating state were handled and include positioning and measurement uncertainty. A reading only at the hotspot centre is insufficient.

---

Exercises

Check your understanding 4

A narrow interference peak occupies only a small part of the required averaging area. Why might its point value be higher than the area-average value?

Show answer

The area average includes the lower values around the peak as well as the peak itself. The point maximum therefore need not equal the comparison quantity. You must still search for the region with the highest required average and perform any separate focusing check.