Wireless & Health

This page is reference material for people who work with wireless equipment. It is not medical advice. It summarizes what regulators and health authorities report; it cannot account for your individual health, and nothing here should be used to diagnose, treat, or rule out any condition. For personal health questions, talk to a qualified clinician. The exposure-reduction steps below are the same optional precautions national agencies publish for people who wish to lower their RF exposure; agencies present them as precautionary, not as remedies for a proven hazard.

All electromagnetic radiation carries energy in proportion to its frequency. The U.S. Federal Communications Commission (FCC) defines radiofrequency fields as radiation from about 3 kHz to 300 GHz. Wi-Fi and most cellular bands sit in the microwave portion of that range: 2.4 GHz Wi-Fi falls in the microwave region, microwave ovens run at 2.45 GHz, and common mobile bands include 900 MHz and 1.8 GHz.

The critical split is between ionizing and non-ionizing radiation. Ionizing radiation (X-rays and gamma rays, at the high-frequency end) carries enough energy per photon to strip electrons from atoms and break chemical bonds, which is the mechanism that can directly damage DNA. RF is non-ionizing: its photons cannot ionize atoms or break bonds, a property of the frequency itself, not of the transmitter power. Visible light and infrared are also non-ionizing; the ionizing threshold lies up in the ultraviolet range, far above any radio or microwave frequency. Because RF cannot ionize, its route to a biological effect is fundamentally different from that of X-rays.

RF exposure is the rate at which the body absorbs RF energy. In the cellular and Wi-Fi bands, where energy is deposited inside tissue, exposure is quantified by the Specific Absorption Rate (SAR): watts of RF energy absorbed per kilogram of tissue (W/kg). SAR is reported either as a whole-body average or as a local, spatial-peak value over a small mass of tissue, the latter being what matters for a device held against the body.

At these frequencies the one established biological mechanism is heating. The FCC, the World Health Organization (WHO), and the International Commission on Non-Ionizing Radiation Protection (ICNIRP) all state that tissue heating is the main interaction between RF fields and the body. This is not a trivial point: high-power RF cooks tissue. A microwave oven heats food at 2.45 GHz, and the FCC notes that high-level RF exposure has long been known to be harmful through rapid heating, with documented effects in laboratory animals including cataracts and temporary sterility. Exposure limits exist precisely because RF at sufficient power is dangerous; the limits are set to keep heating far below any level that could matter.

Limits are anchored to an experimentally observed threshold. The FCC and ICNIRP both use a whole-body SAR of about 4 W/kg as the level above which measurable effects appear in laboratory animals, then apply large reduction factors below it.

In the United States the FCC rules derive from IEEE (ANSI/IEEE C95.1) and NCRP standards. From the 4 W/kg basis:

• Occupational / controlled whole-body average SAR limit: 0.4 W/kg, a factor of 10 below the threshold.
• General population / uncontrolled whole-body average SAR limit: 0.08 W/kg, a factor of 50 below the threshold.
• For portable devices used against the body (phones, tablets), the FCC local SAR limit is 1.6 W/kg averaged over 1 gram of tissue. Every phone legally sold in the United States must meet it.

ICNIRP plays the same role internationally. Its 2020 guidelines cover 100 kHz to 300 GHz, use the same roughly 4 W/kg basis, and apply a reduction factor of 50 for the general public, giving 0.08 W/kg whole-body average. ICNIRP averages local SAR over a 10-gram region (so its local numbers are defined differently from the FCC's 1-gram figure) and, above 6 GHz, switches to an absorbed-power-density restriction because energy is then deposited at the surface rather than deep in tissue. Many countries adopt the ICNIRP limits directly.

The practical point is the size of the margin. Certified consumer Wi-Fi and mobile equipment is type-tested to operate below these limits, which themselves sit 10 to 50 times below the only level at which an effect has been demonstrated. A Wi-Fi access point radiates on the order of a fraction of a watt and, unlike a phone, normally sits away from the body, so exposure from it falls off steeply with distance.

It helps to separate three statements that are easy to blur together.

Established: heating, and the limits that prevent it. The only substantiated effect of RF at consumer frequencies is heating, and the exposure limits keep heating far below any harmful level. WHO and ICNIRP both conclude that, below the limits, no other effect is reliably demonstrated.

Not established, but not disproven: low-level, long-term effects. Below the heating threshold the evidence is, in the FCC's own words, ambiguous and unproven. Crucially, no confirmed adverse effect at typical exposure is not the same as proof of safety. WHO states that no adverse health effect has been causally linked to wireless technologies at exposures within the guidelines, while at the same time calling for more research into long-term effects, and noting the lack of data for mobile-phone use over periods longer than 15 years. Absence of established harm and proof of harmlessness are different claims, and the authorities are careful to make only the first.

A flagged question: a possible cancer association. In 2011 the WHO's International Agency for Research on Cancer (IARC) classified RF electromagnetic fields as Group 2B, "possibly carcinogenic to humans," based on limited evidence of an association between heavy mobile-phone use and glioma. Group 2B is a strength-of-evidence category used when a causal link is considered credible but chance, bias, or confounding cannot be ruled out; it ranks how sure we are, not how dangerous something is, and the group descriptors carry no quantitative risk meaning. It is a large, mixed category that has included agents such as pickled vegetables and gasoline-engine exhaust. The signal IARC weighed appeared mainly in the heaviest users (on the order of 30 minutes a day over years).

The animal and human studies behind the open question are worth knowing, and so are their limits. The U.S. National Toxicology Program (NTP) found "clear evidence" of heart schwannomas in male rats, but at whole-body exposures far higher and longer (about 9 hours a day for life) than human use, and the strongest finding did not appear in female rats or in mice; the FDA says the result should not be applied directly to human phone use. The Ramazzini Institute reported a similar male-rat heart-tumor signal and viewed it as supporting NTP, while ICNIRP judged neither study a reliable basis to change the limits. Interphone, the largest human case-control study, found no overall increased risk, with a weak suggestion of glioma in the highest call-time decile that its own authors said biases prevent interpreting causally. Authorities differ on children: the FDA states current evidence shows no danger to any users including children, while France's ANSES applies precaution specifically for children (smaller bodies, developing tissue) and recommends moderate use, hands-free kits, and limiting calls. That divergence is a policy choice under uncertainty, not a settled fact.

The honest takeaway for a field operator: ordinary Wi-Fi and cellular equipment is non-ionizing and is certified well below internationally harmonized limits that already sit far below the only demonstrated effect threshold; and long-term epidemiology remains an open research area, which is exactly why the limits keep large safety factors and why precautionary steps are reasonable for anyone who wants them.

If you want to lower your RF exposure, the physics points to two levers: distance and power. Exposure falls with the square of distance (double the distance, quarter the exposure), and a phone transmits harder when it has to. The steps national agencies (FCC, FDA, CDC, ANSES) publish all follow from those two facts:

• Put distance between the device and your body. Use speakerphone, wired earbuds, or an air-tube headset for calls instead of holding the phone to your head. Distance is the single most effective lever.
• Text or use data instead of long calls, which keeps the phone away from your head and transmits in short bursts.
• Do not make calls in weak-signal areas if you can avoid it. A phone with poor reception boosts its transmit power to reach the network, raising your exposure.
• Keep the phone off your body when you can. Carry it in a bag rather than a pocket, and do not sleep with it on the pillow. Body-worn SAR compliance assumes a small separation.
• Use airplane mode when you do not need connectivity (overnight, in a drawer). With the radios off, the device does not transmit at all.
• Give Wi-Fi routers some distance. Do not sit with your head next to an access point all day, and prefer wired Ethernet for fixed devices where it is practical. Same inverse-square logic.
• For children, ANSES specifically advises moderate, supervised use and favoring hands-free kits. The FDA does not consider this necessary but offers the same optional steps to anyone who wants them.

These are precautions, not treatments. They cost little and reduce a quantity (absorbed RF) whose long-term effects are not fully characterized; that is the whole case for them.

A Faraday enclosure is a conductive shell that keeps external fields out and internal fields in. When a field reaches the conductor, the metal's free electrons rearrange and flow as surface currents that produce an opposing field, cancelling the field inside. For a time-varying signal the charges keep moving, so the cancellation is continuous. This is why a sealed conductive bag can make a phone effectively vanish from the network: with the field blocked, the phone cannot transmit or receive, so it cannot be reached, tracked, or updated while enclosed.

Two facts explain why thin metal and even mesh work. First, skin depth: RF currents ride in a thin surface layer of the conductor that shrinks as frequency rises (a couple of micrometers in copper at 1 GHz), so a thin foil or coating shields high frequencies well. Second, the aperture rule: a hole leaks a wave only when the hole is a meaningful fraction of the wavelength. An opening shields well when it is much smaller than the wavelength (a common rule of thumb is no larger than about one twentieth of it), and shielding collapses once an opening approaches half a wavelength. A microwave-oven door is the everyday example: its perforated metal screen blocks the 12 cm, 2.45 GHz field while passing visible light, whose wavelength is hundreds of nanometers, far smaller than the holes.

Faraday enclosures have real, legitimate uses: silencing a phone so it cannot be located or reached, blocking a car key fob to defeat relay-amplification theft, and isolating a seized phone in digital forensics so it cannot be remotely wiped (NIST mobile-forensics guidance calls for exactly this RF isolation). They also have real limits. The shell must fully seal; a gap, an open seam, or a weak closure leaks, and independent testing shows wide variation between products and that some bags fail at the higher bands. Simple conductive bags block high-frequency RF well but poorly attenuate low-frequency magnetic fields. A phone sealed inside also burns battery searching for a network at full power. The honest test is empirical: put the phone in, call it and try to locate it; if it rings or shows a location, the bag has failed. And a Faraday bag is a physical guarantee in a way airplane mode is not: airplane mode trusts the operating system to actually stop transmitting, while the bag blocks RF regardless of what the device is doing, which matters given proof-of-concept malware that fakes a powered-off state.

The same radios that connect you also announce you. This is descriptive, for awareness and self-defense; tracking another person without consent is illegal in many places. There are four broad channels.

Cellular. As long as a phone is on and has a SIM, the carrier can place it. Every interaction with a tower generates time-stamped cell-site location information (CSLI); signal strength across several towers triangulates the phone, and a "tower dump" hands investigators every device that touched a given tower in a time window, sweeping in bystanders. In Carpenter v. United States (2018) the U.S. Supreme Court held that acquiring historical CSLI is a Fourth Amendment search generally requiring a warrant, and noted CSLI precision is approaching GPS-level as cells shrink.

Cell-site simulators (IMSI catchers, "Stingrays"). These devices impersonate a cell tower; phones connect to the strongest apparent tower, so nearby phones attach to the simulator and reveal their identifiers (IMSI) and approximate location, with no help from the carrier. Some can force a downgrade to weakly encrypted 2G and intercept more. They are dragnet by nature, pulling in everyone in range, are used by numerous federal and local agencies, and are hard to detect; newer units operate on modern standards.

Wi-Fi. A phone hunting for networks broadcasts probe requests containing its MAC address, and historically the names of remembered networks. Retailers, airports, and transit systems have used MAC-based analytics to follow movement (the FTC's first such case targeted a retail-tracking firm). Phones now randomize their MAC per network to blunt this, but research shows randomization can be partly defeated by fingerprinting and timing. Separately, Apple and Google maintain Wi-Fi positioning databases that map access-point identifiers to locations; 2024 research showed Apple's system could be abused to track access points worldwide, after which mitigations were added.

Bluetooth, Find My, and AirTags. BLE devices broadcast advertisements on open channels. Apple's Find My turns hundreds of millions of Apple devices into a crowd-sourced locator: a lost item beacons a rotating key that any passing iPhone relays back, end-to-end encrypted. The same mechanism enables covert stalking with item trackers, which led Apple and Google to a cross-industry unwanted-tracker detection standard now shipping on iOS and Android. BLE address randomization helps but, like Wi-Fi, has documented weaknesses.

Apps and data brokers. Many apps embed location SDKs that sell precise, timestamped location tied to an advertising ID. The FTC has taken a series of actions against brokers (X-Mode/Outlogic, Kochava, InMarket, Gravy/Venntel) for selling data that could trace visits to clinics, places of worship, and shelters, and has shown how "home at night" plus public records re-identifies a supposedly anonymous device.

Each defense closes a specific channel above:

• Airplane mode, or fully powering off, stops cellular, Wi-Fi, Bluetooth, and GPS transmission. Note that apps may cache location and upload it on reconnect, so a full power-off is the stronger guarantee. There is no way to hide from cellular tracking while the phone is on with a live SIM.
• Turn off Wi-Fi and Bluetooth when not in use, and also disable the separate "scanning" toggles (Android keeps Wi-Fi and Bluetooth scanning options that keep working for location even when the radios look off).
• Keep MAC address randomization on (default on current iOS and Android) to reduce cross-venue Wi-Fi tracking, understanding it is a strong default, not a guarantee.
• Use a Faraday pouch for a hardware guarantee when you need certainty (it blocks every channel at once), remembering it must be fully sealed.
• Run unwanted-tracker detection (built into modern iOS and Android, plus Apple's Tracker Detect app for Android) to catch a BLE tracker that is traveling with you.
• Minimize always-on radios and location permissions: audit per-app location access, reset or disable the advertising ID that brokers use to link data, and for sensitive trips use airplane mode with offline maps.
• Consider a hardened, de-Googled phone (for example GrapheneOS) if your threat model warrants it; such systems add network and sensor permission controls and avoid default background connections.

None of this is about evading lawful process; it is about understanding, and reducing, unwanted commercial and covert tracking.

A common worry is that many radios in one place must add up to harm. The physics says otherwise. RF exposure is a sum of power densities, and each low-power source contributes little. A laptop, phone, and access point each operate far below the limits individually, and their combined exposure anywhere in a room is still a small fraction of the general-public limit, because power density falls with the square of distance from each emitter and each device transmits in short bursts. A dense deployment of access points does not change the per-emitter physics; it mainly improves link quality so clients can transmit at lower power. The relevant comparison is always total exposure against the limit, and in normal indoor environments that total stays well under it. WHO notes that even near typical base stations, measured RF levels are a small fraction of the international limits.

• Spectrum and Regulation for how spectrum is governed and devices are authorized
• RF, Density and Tuning for propagation, power, and how exposure falls with distance
• Glossary for terms such as SAR, EIRP, MAC randomization, and IMSI catcher

Health, exposure, and limits:

• World Health Organization, "Radiation: Electromagnetic fields" (Q&A): https://www.who.int/news-room/questions-and-answers/item/radiation-electromagnetic-fields
• World Health Organization, "Radiation: 5G mobile networks and health" (Q&A): https://www.who.int/news-room/questions-and-answers/item/radiation-5g-mobile-networks-and-health
• World Health Organization, "Electromagnetic fields and public health: mobile phones" (fact sheet): https://www.who.int/news-room/fact-sheets/detail/electromagnetic-fields-and-public-health-mobile-phones
• ICNIRP, "Guidelines for Limiting Exposure to Electromagnetic Fields (100 kHz to 300 GHz)," 2020: https://www.icnirp.org/cms/upload/publications/ICNIRPrfgdl2020.pdf
• ICNIRP, RF EMF overview and FAQ: https://www.icnirp.org/en/frequencies/radiofrequency/index.html
• FCC, "Radio Frequency Safety": https://www.fcc.gov/general/radio-frequency-safety-0
• FCC OET Bulletin 56, "Questions and Answers about Biological Effects and Potential Hazards of Radiofrequency Electromagnetic Fields": https://transition.fcc.gov/Bureaus/Engineering_Technology/Documents/bulletins/oet56/oet56e4.pdf
• FCC, "Wireless Devices and Health Concerns": https://www.fcc.gov/consumers/guides/wireless-devices-and-health-concerns
• IARC (WHO), Press Release No. 208, "IARC classifies Radiofrequency Electromagnetic Fields as possibly carcinogenic to humans," 2011: https://www.iarc.who.int/wp-content/uploads/2018/07/pr208_E.pdf
• IARC Monographs, List of Classifications: https://monographs.iarc.who.int/list-of-classifications
• US National Toxicology Program / NIEHS, "Cell Phone Radio Frequency Radiation": https://www.niehs.nih.gov/health/topics/agents/cellphones
• FDA, "Scientific Evidence for Cell Phone Safety": https://www.fda.gov/radiation-emitting-products/cell-phones/scientific-evidence-cell-phone-safety
• FDA, "Reducing Radio Frequency Exposure from Cell Phones": https://www.fda.gov/radiation-emitting-products/cell-phones/reducing-radio-frequency-exposure-cell-phones
• CDC, "Facts About Cell Phones and Your Health": https://www.cdc.gov/radiation-health/data-research/facts-stats/cell-phones.html
• ANSES, "Exposure of children to radiofrequencies": https://www.anses.fr/en/content/exposure-children-radiofrequencies-call-moderate-and-supervised-use-wireless-technologies

Faraday shielding:

• "Electromagnetic shielding," overview of shielding theory: https://en.wikipedia.org/wiki/Electromagnetic_shielding
• LearnEMC, "Introduction to Practical Electromagnetic Shielding": https://learnemc.com/practical-em-shielding
• Academy of EMC, "Shielding" (aperture and wavelength, shielding effectiveness): https://www.academyofemc.com/shielding
• NIST SP 800-101 Rev. 1, "Guidelines on Mobile Device Forensics" (RF isolation): https://csrc.nist.gov/pubs/sp/800/101/r1/final
• Matt Blaze, "Testing Phone-Sized Faraday Bags" (independent measurements): https://www.mattblaze.org/blog/faraday/

Tracking and defenses:

• Carpenter v. United States, U.S. Supreme Court slip opinion (No. 16-402), 2018: https://www.supremecourt.gov/opinions/17pdf/16-402_h315.pdf
• EFF Surveillance Self-Defense, "Mobile Phones: Location Tracking": https://ssd.eff.org/module/mobile-phones-location-tracking
• EFF Street-Level Surveillance, "Cell-Site Simulators / IMSI Catchers": https://sls.eff.org/technologies/cell-site-simulators-imsi-catchers
• ACLU, "Stingray Tracking Devices": https://www.aclu.org/issues/privacy-technology/surveillance-technologies/stingray-tracking-devices
• "A Study of MAC Address Randomization in Mobile Devices and When It Fails": https://arxiv.org/abs/1703.02874
• Rye and Levin, "Surveilling the Masses with Wi-Fi-Based Positioning Systems," IEEE S&P 2024: https://arxiv.org/abs/2405.14975
• Apple Platform Security, "Find My Security": https://support.apple.com/guide/security/find-my-security-sec6cbc80fd0/web
• Apple and Google, "Detecting Unwanted Location Trackers" (IETF draft): https://datatracker.ietf.org/doc/html/draft-ledvina-apple-google-unwanted-trackers-00
• Becker, Li, and Starobinski, "Tracking Anonymized Bluetooth Devices," PoPETs 2019: https://petsymposium.org/popets/2019/popets-2019-0036.pdf
• FTC, "FTC Order Prohibits Data Broker X-Mode Social and Outlogic from Selling Sensitive Location Data," 2024: https://www.ftc.gov/news-events/news/press-releases/2024/01/ftc-order-prohibits-data-broker-x-mode-social-outlogic-selling-sensitive-location-data
• FTC, "FTC Sues Kochava for Selling Data that Tracks People," 2022: https://www.ftc.gov/news-events/news/press-releases/2022/08/ftc-sues-kochava-selling-data-tracks-people-reproductive-health-clinics-places-worship-other
• GrapheneOS, "Features Overview": https://grapheneos.org/features