Electronics Guide

Organization and Mandate

ICNIRP was established in 1992 as a successor to the International Non-Ionizing Radiation Committee of the International Radiation Protection Association. Key characteristics include:

• Scientific independence: ICNIRP is not affiliated with any government or industry. Commission members are selected based on scientific expertise and independence from commercial interests.
• Comprehensive scope: ICNIRP addresses all non-ionizing radiation, including static fields, extremely low frequency fields, radiofrequency fields, optical radiation, and ultrasound.
• Evidence-based approach: Guidelines are based on systematic review of peer-reviewed scientific literature, with emphasis on established adverse health effects rather than unconfirmed or controversial findings.
• Regular updates: ICNIRP periodically revises guidelines as new evidence becomes available. The most recent comprehensive RF guidelines were published in 2020, replacing the 1998 guidelines.

Basic Restrictions and Reference Levels

ICNIRP guidelines employ a two-tier approach:

• Basic restrictions: Limits on the quantities directly related to established health effects. For RF fields, the basic restriction is Specific Absorption Rate (SAR), which directly determines tissue heating; above 6 GHz, where absorption is confined to the surface, absorbed power density takes over as the local restriction. For low-frequency fields, basic restrictions are expressed as induced electric field in the body, which relates to nerve and muscle stimulation.
• Reference levels: Derived quantities that can be measured or calculated in the environment outside the body. Reference levels are expressed as incident field strengths (E-field in V/m, H-field in A/m) or incident power density (W/m²). ICNIRP derives them from the basic restrictions under worst-case coupling assumptions, so compliance with a reference level normally demonstrates compliance with the corresponding basic restriction. ICNIRP stops short of calling this a guarantee, and states that a few exposure scenarios at the reference levels could still exceed a basic restriction.

Compliance with either tier suffices, and demonstrating both is not required. When reference levels are exceeded, compliance can still be shown by demonstrating through dosimetric assessment that the basic restrictions are met. In one case the choice disappears: above 2 GHz, inside the reactive near-field zone, ICNIRP treats the incident-field quantities as inadequate evidence, so the basic restrictions must be assessed directly.

Frequency-Dependent Limits

ICNIRP guidelines specify different limits for different frequency ranges, reflecting the frequency-dependent nature of biological effects:

• 1 Hz to 100 kHz: Limits protect against nerve and muscle stimulation from induced currents. The basic restriction is induced electric field in the central nervous system (CNS) or peripheral nervous system (PNS). Reference levels decrease with increasing frequency in this range.
• 100 kHz to 10 MHz: A transition region where both thermal and stimulation effects must be considered. Limits are set to protect against both mechanisms.
• 10 MHz to 300 GHz: Limits protect against excessive tissue heating. Under the 2020 ICNIRP guidelines the whole-body average SAR basic restriction is 0.08 W/kg for the general public and 0.4 W/kg for occupational exposure, a factor of five apart, and it applies unchanged across the entire range to 300 GHz. Only the local restriction changes quantity at 6 GHz. From 100 kHz to 6 GHz it is peak spatial-average SAR: 2 W/kg (public) and 10 W/kg (occupational) for the head and torso, and 4 and 20 W/kg for the limbs. Above 6 GHz it becomes absorbed power density: 20 W/m² (public) and 100 W/m² (occupational). The corresponding whole-body reference level above 2 GHz is an incident power density of 10 W/m² for the public and 50 W/m² for occupational exposure.

2020 Updates

The 2020 ICNIRP guidelines incorporated several significant changes:

• Higher frequency coverage: Extended detailed treatment to 300 GHz, addressing 5G and other emerging technologies.
• Refined local SAR averaging: Changed the local, or peak spatial-average, SAR averaging volume from any 10 g of contiguous tissue (1998) to a 10 g cubic mass (2020). The numeric SAR restrictions themselves did not move. Both editions set 0.4 and 0.08 W/kg whole-body average, 10 and 2 W/kg local head and torso, and 20 and 4 W/kg local limb, for occupational and general public exposure respectively.
• A new quantity above 6 GHz: The 1998 guidelines restricted power density above 10 GHz, averaged over any 20 cm² of exposed area and over a frequency-dependent period of 68/f^1.05 minutes, with f in gigahertz. The 2020 guidelines moved the transition down to 6 GHz, replaced that quantity with absorbed power density inside the body surface, and fixed the averaging at a square 4 cm² of body surface and a flat 6 minutes. Above 30 GHz an additional constraint applies over a square 1 cm², set at twice the 4 cm² value.
• Longer whole-body averaging time: Extended the averaging time for the whole-body average SAR restriction from 6 minutes (1998) to 30 minutes (2020), better matching the time taken for the body core temperature to rise.
• Temperature-based approach: The guidelines more explicitly connected basic restrictions to temperature rise, the actual parameter of concern for thermal effects.
• Brief-exposure provisions: Added restrictions for intervals shorter than the 6-minute averaging time, expressed as specific energy absorption in kJ/kg below 6 GHz and as absorbed energy density in kJ/m² above it, to limit transient heating from brief high-level exposure.
• Local exposure limits: More detailed treatment of localized exposure to small areas of the body, particularly relevant for devices held close to the body.

IEEE C95 Series

The primary IEEE exposure standards are in the C95 series:

• IEEE C95.1: Safety levels with respect to human exposure to electric, magnetic, and electromagnetic fields, 0 Hz to 300 GHz. This is the main exposure limit standard. The 2019 revision (C95.1-2019) consolidated the low-frequency limits formerly in C95.6 with the radiofrequency limits, producing a single standard spanning the full 0 Hz to 300 GHz range.
• IEEE C95.2: Radio-frequency energy and current-flow symbols (safety signage).
• IEEE C95.3: Recommended practice for measurements and computations of electric, magnetic, and electromagnetic fields. Provides measurement methodology.
• IEEE C95.4: Recommended practice for determining safe distances from radio frequency transmitting antennas when using electric blasting caps.
• IEEE C95.6: Safety levels with respect to human exposure to electromagnetic fields, 0 to 3 kHz. The 2002 edition addressed low-frequency exposure separately; its limits were subsequently merged into C95.1-2019.
• IEEE C95.7: Recommended practice for radio frequency safety programs. Addresses implementation of RF safety in workplaces.

Differences from ICNIRP

The most recent revisions, IEEE C95.1-2019 and ICNIRP 2020, were developed in close coordination and are now substantially harmonized. Both share the same whole-body average SAR basic restrictions (0.4 W/kg occupational, 0.08 W/kg general public), the same local SAR limits (10 and 2 W/kg for the head and torso; 20 and 4 W/kg for the limbs), a 10-gram cubic averaging mass for local SAR, a 30-minute whole-body averaging time, and the same transition from SAR to absorbed power density above 6 GHz. Some differences nonetheless remain:

• Tier terminology: IEEE C95.1 sets limits for persons in "restricted" environments, who are aware of their exposure and able to exercise control (broadly comparable to ICNIRP occupational exposure), and persons in "unrestricted" environments (broadly comparable to ICNIRP general public). The two frameworks use different names and supporting-program concepts for what are essentially two equivalent tiers.
• Reference levels and measurement detail: The derived external-field reference levels (which IEEE terms exposure reference levels) and the recommended spatial- and temporal-averaging procedures differ in some particulars, even where the underlying basic restrictions agree.
• Averaging mass in older editions: Before the 2019 harmonization, IEEE C95.1 averaged local SAR over 1 gram of tissue rather than the 10 grams used by ICNIRP, and the United States FCC rules still do. The two figures are not comparable as bare numbers. The FCC limit for general population exposure is 1.6 W/kg averaged over any 1 gram of tissue in the shape of a cube; the ICNIRP 2020 and IEEE C95.1-2019 limit is 2 W/kg averaged over a 10 g cube. A 1-gram average resolves smaller hot spots and returns a higher value for the same exposure, so the lower United States number is the more demanding of the two in practice. The 2019 revision of C95.1 adopted the 10-gram cubic mass, while the FCC limits retain the 1-gram convention for the head and trunk and use 10 grams for the extremities.

Standard Development Process

IEEE standards are developed through a consensus process:

• Working groups: Technical experts from academia, industry, government, and other sectors participate in developing draft standards.
• Balloting: Draft standards are balloted by a broad group of stakeholders. Negative votes with technical comments must be addressed.
• Public comment: Draft standards are available for public comment before final approval.
• Regular revision: Standards are reviewed and revised periodically (typically every 5 to 10 years) to incorporate new scientific knowledge.
• Harmonization efforts: IEEE and ICNIRP have ongoing communication to understand technical differences and move toward greater harmonization where scientifically justified.

European Union

The EU addresses electromagnetic exposure through several instruments:

• Council Recommendation 1999/519/EC: Adopted 12 July 1999, it recommends that member states apply ICNIRP basic restrictions and reference levels to general public exposure from 0 Hz to 300 GHz. Its annexes reproduce the ICNIRP 1998 guidelines, and the Council has not updated it to the 2020 revision, so the European public reference levels still rest on the 1998 figures. A recommendation does not bind member states, but it shapes national legislation and harmonized product standards.
• Directive 2013/35/EU: The twentieth individual directive under the framework Directive 89/391/EEC, adopted 26 June 2013 to replace Directive 2004/40/EC, with a transposition deadline of 1 July 2016. It establishes legally binding requirements for occupational exposure: employers must assess electromagnetic field risks and implement protective measures. Its limits follow ICNIRP recommendations but use their own vocabulary, setting exposure limit values, divided into health effects and sensory effects, together with lower action levels that permit a simplified demonstration of compliance.
• Radio Equipment Directive (2014/53/EU): Requires that radio equipment not expose users to excessive electromagnetic fields. Compliance is demonstrated through harmonized standards that reference ICNIRP limits, among them EN IEC/IEEE 62209-1528 for hand-held and body-worn devices, EN IEC 62232 for base stations, and EN 50663 for low-power equipment.
• Member state implementation: Individual EU countries may have additional requirements. Some countries apply lower limits in specific situations (near schools, hospitals, etc.) based on precautionary approaches.

United States

US electromagnetic exposure requirements come from multiple agencies:

• FCC (Federal Communications Commission): Regulates RF exposure from communications devices and transmitters under 47 CFR 1.1307, 1.1310, 2.1091, and 2.1093. The SAR rules cover 100 kHz to 6 GHz. For general population, or uncontrolled, exposure the limits are 0.08 W/kg whole-body average and a peak spatial-average SAR of 1.6 W/kg over any 1 gram of tissue in the shape of a cube, averaged over a period not exceeding 30 minutes. For occupational, or controlled, exposure they are 0.4 W/kg whole-body average and 8 W/kg over 1 gram, averaged over a period not exceeding 6 minutes. Extremities, which the rules take to include the hands, wrists, feet, ankles, and pinnae, are allowed 4 and 20 W/kg over 10 grams. The FCC drew the SAR criteria from ANSI/IEEE C95.1-1992 and NCRP Report No. 86 rather than from any current edition, which is why they differ from present IEEE and ICNIRP values.
• OSHA (Occupational Safety and Health Administration): Has occupational exposure limits but generally defers to other standards. Specific requirements exist for some workplace hazards involving non-ionizing radiation.
• FDA (Food and Drug Administration): Regulates medical devices including those that emit electromagnetic fields. Works with FCC on wireless device safety.
• State regulations: Some states have additional requirements, particularly for occupational exposure or specific facility types.

Other Major Jurisdictions

Other significant regulatory frameworks include:

• China: GB 8702-2014, "Controlling Limits for Electromagnetic Environment," sets limits from 1 Hz to 300 GHz, drawing on ICNIRP guidelines with some differences. Its scope is public exposure in the ambient electromagnetic environment. Occupational exposure and exposure during medical diagnosis or treatment fall outside it and are covered by separate standards.
• Japan: Radio Radiation Protection Guidelines are set by the Ministry of Internal Affairs and Communications. Limits generally follow ICNIRP guidelines.
• Canada: Health Canada's Safety Code 6 establishes RF exposure limits. The limits are broadly consistent with ICNIRP, with specific Canadian considerations. Innovation, Science and Economic Development Canada (ISED, formerly Industry Canada) enforces compliance for radio equipment.
• Australia: ARPANSA (Australian Radiation Protection and Nuclear Safety Agency) publishes Radiation Protection Standard for Limiting Exposure to Radiofrequency Fields (RPS S-1), adopting ICNIRP guidelines.
• International Telecommunication Union: ITU-T K.52 and related recommendations address human exposure from telecommunications installations, referencing ICNIRP guidelines.

Rationale for Higher Limits

Occupational limits are typically higher than general public limits for several reasons:

• Informed exposure: Workers are (or should be) aware of electromagnetic hazards and trained in protective measures. They can take appropriate precautions when working in high-field areas.
• Controlled conditions: Workplace exposures occur under controlled conditions with access restrictions, monitoring, and safety programs. Uncontrolled public access does not occur.
• Working-age adults: Occupational limits apply to healthy working-age adults, not to potentially more susceptible groups (children, elderly, people with medical conditions) who may be present in the general public.
• Limited duration: Occupational exposure typically occurs during working hours only, while general public exposure may be continuous.

ICNIRP occupational basic restrictions (SAR) are 5 times higher than general public restrictions, corresponding to a factor of approximately 2.2 in field strength reference levels.

Occupational Categories

Workers with significant electromagnetic exposure include:

• Broadcast and telecommunications: Workers maintaining transmitting antennas, climbing towers, or working near high-power transmitters may experience significant RF exposure.
• Industrial RF applications: Operators of RF heat sealers, induction heaters, dielectric heaters, and similar industrial equipment can be exposed to high field levels.
• Medical applications: Healthcare workers using MRI, diathermy, electrosurgery, and other medical equipment may be exposed to elevated fields.
• Electrical utilities: Workers near high-voltage power lines and in substations experience ELF electric and magnetic fields.
• Military and security: Operation of radar systems, electronic warfare equipment, and security screening devices can create significant exposures.
• Research and testing: Personnel in EMC test laboratories, antenna test ranges, and research facilities may work near high-power electromagnetic sources.

Occupational Safety Programs

Effective occupational electromagnetic safety requires comprehensive programs:

• Hazard identification: All workplace electromagnetic sources must be identified and characterized. This includes fixed installations, portable equipment, and transient sources.
• Exposure assessment: Measurements or calculations determine actual exposure levels and compare them to limits. Assessment should address both routine operations and worst-case scenarios.
• Control measures: Engineering controls (shielding, interlocks, reduced power) and administrative controls (time limits, work procedures, access restrictions) reduce exposure where necessary.
• Training: Workers must understand electromagnetic hazards, recognize warning signs, and know protective measures. Training should be documented and periodically refreshed.
• Personal protective equipment: While less common than for other hazards, RF-protective clothing and eyewear may be appropriate in specific situations.
• Medical surveillance: Some programs include medical monitoring for workers with significant exposure, though the value of routine surveillance for electromagnetic exposure is debated.

Protection of Sensitive Groups

General public limits must protect populations that may be more vulnerable:

• Children: Smaller body size may result in different SAR distribution. Developing tissues might respond differently to exposure. Conservative limits accommodate these possibilities.
• Pregnant women and fetuses: Prenatal exposure concerns have driven research on developmental effects. Current limits include safety factors intended to protect the developing fetus.
• Elderly persons: Reduced thermoregulatory capacity may decrease tolerance for thermal loads from RF exposure.
• People with medical devices: Pacemakers and other implants may be affected by electromagnetic fields at levels below those causing direct health effects. While device immunity is addressed separately, conservative public limits provide additional margin.
• People with medical conditions: Various health conditions might affect susceptibility to electromagnetic fields, though specific vulnerabilities are not well characterized.

Continuous Exposure Considerations

Unlike occupational exposure, general public exposure may be continuous:

• Residential exposure: People living near broadcasting antennas, power lines, or cellular base stations may be exposed 24 hours per day.
• Environmental sources: Background electromagnetic fields from multiple sources create baseline exposure that is essentially continuous.
• Personal devices: Mobile phones, wireless devices, and wearables create exposure during much of the day.
• Time averaging implications: Standards specify time averaging periods for compliance assessment. For truly continuous exposure, the average equals the peak. For intermittent exposure, higher instantaneous levels may be permissible if the time average remains compliant.

Public Exposure Scenarios

Common sources of general public electromagnetic exposure include:

• Mobile phones and wireless devices: The dominant source of RF exposure for most people, particularly to the head and body. Subject to specific SAR limits and compliance testing.
• Cellular base stations: Despite public concern, exposure levels at publicly accessible locations are typically far below limits (often less than 1% of reference levels).
• Broadcasting transmitters: AM, FM, and television transmitters can create elevated fields near the tower, but public access is normally restricted.
• Wi-Fi and wireless networks: Low power levels result in exposure far below limits except very close to the device.
• Power lines and electrical infrastructure: Create ELF electric and magnetic fields. Fields decrease rapidly with distance and are typically well below limits at normal distances.
• Home appliances: Various appliances produce localized electromagnetic fields, but exposure is typically brief and well below limits.

Summation Rules

When exposure occurs at multiple frequencies, the effects may be additive. Standards specify summation formulas:

Thermal effects (100 kHz to 300 GHz): For frequencies where thermal effects dominate, the basic restrictions are dosimetric quantities (SAR and absorbed power density) that are proportional to absorbed power, so their contributions add linearly:

Sum over all frequencies of (SAR_i / SAR_limit) must be less than or equal to 1

When compliance is evaluated against the external-field reference levels instead, the incident electric- and magnetic-field-strength ratios are squared before summing, because absorbed power is proportional to the square of the field:

Sum over all frequencies of (E_inc,i / E_inc,limit)² must be less than or equal to 1

Incident power density ratios, being power quantities, are summed linearly rather than squared.

Stimulation effects (1 Hz to 10 MHz): For low frequencies, where nerve and muscle stimulation is the concern, the basic restriction is the electric field induced inside the body, and those contributions are summed without squaring:

Sum over all frequencies of (E_ind,i / E_ind,limit) must be less than or equal to 1

The two expressions use different quantities and should not be confused. E_inc is the incident electric field strength, measured outside the body and compared against a reference level. E_ind is the electric field induced in tissue, a basic restriction that ICNIRP evaluates as a root-mean-square value over a 2 mm cube of contiguous tissue. The summation is linear here because stimulation depends on field strength rather than on absorbed power. ICNIRP applies the same linear form to the low-frequency external reference levels, summing the incident electric field ratios and the incident magnetic field ratios separately.

Transition region (100 kHz to 10 MHz): The two mechanisms overlap in this band. The thermal restrictions begin at 100 kHz and ICNIRP carries the induced-field restrictions up to 10 MHz, so throughout the overlap both summations must be satisfied.

Practical Implementation

Applying multiple frequency assessment in practice involves:

• Frequency identification: All significant exposure frequencies must be identified. This may require spectrum analysis or detailed knowledge of source characteristics.
• Individual assessment: Exposure at each frequency is assessed against the corresponding limit.
• Summation calculation: The appropriate summation formula is applied. Frequencies are grouped according to the applicable effect mechanism.
• Conservative assumptions: When exact exposure levels are uncertain, worst-case assumptions should be used. Phase relationships between sources are typically unknown and assumed unfavorable.

Common Multi-Frequency Scenarios

Typical situations involving multiple frequency exposure:

• Near cellular base stations: Multiple cellular bands (700 MHz, 850 MHz, 1900 MHz, 2100 MHz, etc.) plus Wi-Fi and other services may be present. Each contribution is typically small, and the sum remains well below limits.
• In vehicles: Multiple wireless devices (cellular, Wi-Fi, Bluetooth, keyless entry, tire pressure monitors) operate simultaneously. Levels are generally low.
• Industrial environments: Workers may be exposed to power frequency fields from equipment plus RF from wireless communications and process control systems.
• Near broadcast sites: Multiple broadcast transmitters (AM, FM, TV, digital) operating at different frequencies may be co-located.

Averaging Periods

Different averaging periods apply depending on frequency and effect mechanism:

• RF thermal effects: Tissue heating is a thermally integrated process with time constants on the order of minutes, so limits are averaged over time. Under ICNIRP 2020 (and IEEE C95.1-2019), the whole-body average SAR is averaged over 30 minutes, while every local restriction uses a 6-minute averaging time, SAR below 6 GHz and absorbed power density above it. The 2020 revision dropped the frequency-dependent averaging period of the 1998 guidelines, which shortened as 68/f^1.05 minutes with f in gigahertz, and constrained short exposures with a separate rule instead.
• Low-frequency stimulation: Much shorter averaging periods apply (typically single-pulse or single-cycle assessment) because nerve and muscle stimulation depends on instantaneous field values, not long-term averages.
• Brief intervals: Averaging over a fixed window would otherwise permit an arbitrarily intense burst inside it, so ICNIRP 2020 restricts every interval shorter than 6 minutes as well. Above 400 MHz the restricted quantity is the specific energy absorption in a 10 g cube. For occupational exposure of the head and torso the restriction is 3.6[0.05 + 0.95√(t/360)] kJ/kg, with t the interval in seconds, and the general public figure is one fifth of that. At t = 360 s the expression returns 3.6 kJ/kg, which is exactly the 10 W/kg six-minute restriction; at t = 3.6 s it returns 0.52 kJ/kg, an average of roughly 145 W/kg over that interval. A short burst may therefore run well above the six-minute figure, but its allowance grows only as the square root of the interval.

Application of Time Averaging

Time averaging is applied differently for various exposure scenarios:

• Continuous exposure: If exposure is constant, the time-averaged value equals the instantaneous value, and averaging provides no relief.
• Intermittent exposure: If exposure occurs only part of the time (for example, a radar with a rotating antenna), the time-averaged exposure is less than the peak. Higher peak levels may be permissible.
• Varying exposure: If exposure levels vary during the averaging period (for example, due to changing distance from a source), the time-weighted average applies.
• Occupational exposure patterns: Work activities may involve brief high exposures interspersed with lower exposure periods. Time averaging accommodates this pattern if the average remains compliant.

Peak Limits

Time averaging is constrained by peak limits that prevent very high brief exposures:

• ICNIRP approach: The 2020 guidelines constrain peaks in two separate ways. For intervals shorter than the 6-minute window they restrict the energy absorbed, as described above, which caps how far a short burst may exceed the averaged figure. Independently, from 100 kHz to 10 MHz they set peak-value reference levels on the incident fields: 170 V/m and 80 A/m for occupational exposure, and 83 V/m and 21 A/m for the general public. Above 10 MHz, where nerve stimulation is no longer the mechanism of concern, ICNIRP sets no peak instantaneous reference level.
• IEEE approach: Specifies ceiling limits that cannot be exceeded regardless of averaging. For example, SAR values cannot exceed ceiling values even if time-averaged SAR would be compliant.
• Pulsed fields: For pulsed sources (radar, pulsed industrial equipment), peak field strength or SAR during pulses may be much higher than the average. Peak limits ensure that individual pulses do not cause transient adverse effects.

Whole-Body Averaging

For compliance with whole-body exposure limits:

• Field measurement locations: Standards typically specify measuring at multiple heights corresponding to the human body (feet, knees, waist, chest, head) and averaging the results.
• SAR averaging: Whole-body average SAR is the total power absorbed divided by body mass. This inherently involves spatial averaging over the entire body.
• Non-uniform fields: When fields vary significantly over the body dimensions, point measurements must be combined appropriately. Simple averaging may not capture worst-case exposure if fields peak at specific body locations.

Localized Exposure Averaging

For devices used close to the body, localized SAR limits apply:

• SAR averaging mass: ICNIRP 2020 and IEEE C95.1-2019 both specify local SAR averaged over any 10 grams of tissue in the shape of a cube. The United States FCC limits, derived from an earlier IEEE edition, instead use a 1-gram averaging mass for the head and trunk. The smaller mass resolves more localized peaks and generally yields a higher reported SAR for the same exposure.
• Practical implications: A smaller averaging mass captures smaller hot spots and returns higher values, so a handset that meets the 1.6 W/kg figure over 1 gram will in practice also meet the 2 W/kg figure over 10 grams, while the converse does not follow. Manufacturers selling into both markets consequently design to the United States figure, and product documentation commonly reports a 1-gram and a 10-gram value side by side.
• Location dependence: For mobile phones and similar devices, SAR is measured at the location of highest absorption, which varies with device design and use position.

Area Averaging for High Frequencies

At frequencies where absorption is superficial (above approximately 6 GHz), power density rather than SAR becomes the primary metric:

• Averaging area: Absorbed power density is averaged over a square area of the body surface. ICNIRP 2020 uses 4 cm² from above 6 GHz to 300 GHz, and adds a 1 cm² average above 30 GHz, restricted to twice the 4 cm² value, to capture the smaller and more focused absorption spots produced at millimeter-wave frequencies. The 1998 guidelines had no absorbed power density restriction at all; they restricted power density over any 20 cm² of exposed area, and only above 10 GHz.
• Uniform illumination assumption: Reference levels assume relatively uniform exposure over the averaging area. Non-uniform beams may require detailed dosimetric assessment.
• Scanning beams: 5G and other systems using beamforming create spatially varying exposure that must be appropriately averaged over time and space.

Field Measurement Methods

External field measurements assess compliance with reference levels:

• Instrumentation: Broadband field meters measure the total field across a frequency range. Frequency-selective measurements (using spectrum analyzers) characterize individual sources. Both types have applications depending on the assessment scenario.
• Probe types: Electric field probes (typically dipole or monopole elements) and magnetic field probes (typically loop elements) measure the respective field components. Isotropic probes with three orthogonal elements measure the total field regardless of orientation.
• Calibration: All measurement equipment must be calibrated by accredited laboratories. Calibration should be traceable to national standards and should cover the frequency and amplitude ranges of interest.
• Measurement uncertainty: All measurements have uncertainty that must be characterized and reported. Uncertainty sources include calibration uncertainty, environmental effects, positioning errors, and field non-uniformity.

SAR Measurement Methods

For devices used close to the body, SAR is measured under standardized procedures. IEC/IEEE 62209-1528 governs hand-held and body-worn devices from 4 MHz to 10 GHz, superseding the earlier IEC 62209-1, IEC 62209-2, and IEEE 1528 documents. IEC 62232 covers assessment in the vicinity of radiocommunication base stations, and IEEE C95.3 gives the general practice for field measurement and computation.

• Phantom models: Standardized tissue-simulating phantoms (such as the SAM head phantom for mobile phones) provide reproducible measurement conditions. Phantom dielectric properties are specified at test frequencies.
• Probe scanning: An electric field probe scans through the phantom liquid to measure internal field distribution. SAR is calculated from field values and tissue properties.
• Temperature rise methods: Some systems measure temperature rise in phantoms using fiber optic or other non-metallic thermometers. SAR is calculated from the initial rate of temperature increase.
• Test positions: Devices are tested in multiple positions representing different use cases (phone against head, phone in pocket, tablet on lap, etc.). The highest SAR value determines compliance.

Computational Assessment

When direct measurement is impractical, computational methods assess exposure:

• Numerical methods: Finite-Difference Time-Domain (FDTD), Finite Element Method (FEM), and other computational electromagnetics techniques calculate fields in and around the body.
• Anatomical models: Accurate human body models with detailed tissue geometry and frequency-dependent dielectric properties are essential. Standardized models enable reproducible calculations.
• Validation: Computational methods should be validated against measurements in standardized configurations. Agreement between calculation and measurement provides confidence in results.
• Uncertainty estimation: Computational uncertainty arises from discretization, material property values, model geometry, and numerical convergence. Uncertainty should be estimated and reported.

Documentation and Reporting

Compliance assessments require thorough documentation:

• Test conditions: All relevant parameters must be recorded, including source characteristics, measurement location, equipment used, environmental conditions, and personnel present.
• Results presentation: Results should clearly state measured or calculated values, applicable limits, and compliance determination. Spatial and temporal averaging should be explicitly documented.
• Uncertainty statement: The reported measurement or calculation uncertainty allows proper interpretation of results, especially for values near limits.
• Repeatability: Documentation should be sufficient to allow another competent party to reproduce the assessment.