The Electromagnetic Spectrum Explained: From Radio Waves to Gamma Rays

The electromagnetic spectrum is the full range of electromagnetic radiation, from radio waves kilometres long to gamma rays smaller than an atomic nucleus. Every part of it is the same thing, coupled electric and magnetic fields moving at the speed of light. Only frequency changes, and with it wavelength (c = f times lambda) and photon energy (E = h times f).

This post sits beside our Maxwell's equations explainer, which shows where electromagnetic waves come from, and our Fourier transform explainer, which shows how a signal is split into frequencies. Here we walk the whole spectrum from end to end: what each band is, how it is made and detected, why the air above us lets some bands through and stops others, who decides which radio frequencies you may use, where everyday devices sit, and what "ionizing" actually means. Along the way there are worked examples with real numbers, a small calculator you can run, and Linux commands that show which part of the spectrum your own Wi-Fi card is using right now.

Key takeaways

• Radio waves, microwaves, infrared, visible light, ultraviolet, X-rays and gamma rays are one phenomenon. NASA describes all of them as light; the band names are human conventions, not physical boundaries.
• Three numbers describe any point on the spectrum: frequency, wavelength and photon energy. The speed of light (exactly 299,792,458 m/s) links frequency and wavelength, and the Planck constant (exactly 6.62607015 times 10 to the minus 34 J/Hz) links frequency and energy.
• A 2.4 GHz Wi-Fi signal has a wavelength of about 12.5 cm; 5.8 GHz is about 5.2 cm. A green photon carries about 2.25 eV; a 0.1 nm X-ray photon carries about 12.4 keV, roughly 1.25 billion times the energy of a 2.4 GHz photon.
• The atmosphere is opaque to most gamma rays, X-rays and short ultraviolet, partly opaque to infrared, and largely transparent to visible light and much of the radio range. That is why X-ray and gamma-ray telescopes fly in space and why satellites talk to the ground in microwave bands.
• Radio spectrum use is governed by the ITU Radio Regulations, an international treaty revised at World Radiocommunication Conferences, and by national tables. In the US, the FCC handles non-federal use and NTIA handles federal use; only 8.3 kHz to 275 GHz is allocated.
• Wi-Fi works without a licence because rules such as FCC Part 15 permit low power devices in shared bands like 2400 to 2483.5 MHz, 5725 to 5850 MHz and, since 2020, a 1,200 MHz block at 6 GHz.
• Ionizing radiation (gamma rays, X-rays and some ultraviolet) can knock electrons out of atoms. Radio and microwave exposure limits address heating, not ionization; the FCC public limit for phones is a SAR of 1.6 W/kg averaged over 1 gram of tissue.

What electromagnetic radiation is

Changing fields that carry themselves

Charges at rest make electric fields. Charges in steady motion make magnetic fields. When either field changes, it produces the other. NASA's introduction puts it simply: a changing magnetic field induces a changing electric field and the reverse, and the linked pair forms an electromagnetic wave that, unlike sound, needs no medium and crosses the vacuum of space. James Clerk Maxwell wrote down that relationship in the 1860s and 1870s, and his equations predicted a wave whose speed matched the measured speed of light. Heinrich Hertz then produced and detected radio waves in the late 1880s with spark gaps, and showed they behaved like light. The unit of frequency, one cycle per second, carries his name. Our Maxwell's equations explainer works through the four equations themselves; here we only need the result: a wave of coupled fields that travels at c.

The same radiation also behaves like a stream of particles. NASA describes light as discrete packets of energy called photons that carry momentum, have no mass and travel at the speed of light. Which behaviour you see depends on the instrument. A diffraction grating that spreads light into a spectrum is using its wave nature. A digital camera sensor, where individual photons free electrons that are counted, is using its particle nature.

Frequency, wavelength and photon energy

Every electromagnetic wave can be described by three linked quantities.

Frequency (f) is how many wave crests pass a fixed point each second, measured in hertz. A wave with two cycles per second has a frequency of 2 Hz.

Wavelength (lambda) is the distance from one crest to the next, measured in metres. NASA notes that the shortest wavelengths are fractions of the size of an atom, while the longest studied can be larger than the diameter of the Earth.

Photon energy (E) is the energy carried by one photon, measured in joules or, more conveniently, in electron volts (eV). One electron volt is the energy an electron gains moving through a potential difference of one volt.

Two equations tie them together:

c = f × λ        (speed of light = frequency × wavelength)
E = h × f        (photon energy = Planck constant × frequency)
E = h × c / λ    (combine the two)

The constants are not measured approximations any more. NIST's CODATA pages list both as exact:

The elementary charge matters because it converts between joules and electron volts: 1 eV is 1.602176634 × 10^-19 J. Multiply h (in eV per hertz) by c and you get a constant that every spectroscopist memorises:

h × c = 4.135667696e-15 eV·s × 299792458 m/s
      = 1.23984198e-6 eV·m
      = 1239.84 eV·nm

So the energy of a photon in electron volts is 1239.84 divided by its wavelength in nanometres. That single shortcut does most of the work in the worked examples below.

Why scientists describe each band differently

NASA points out a convention that confuses many readers: radio and microwaves are usually described by frequency (hertz), infrared and visible light by wavelength (metres, micrometres, nanometres), and X-rays and gamma rays by energy (electron volts). It is the same wave described in whichever unit keeps the numbers readable. A Wi-Fi engineer says "5 GHz", an astronomer says "21 centimetre line", an optical engineer says "1550 nm" and a radiologist says "120 kV tube". They are all pointing at different places on the same axis.

The bands, from radio to gamma rays

The boundaries are conventions

Nothing physical happens at the line between "microwave" and "infrared" or between "ultraviolet" and "X-ray". The names come from the history of how each band was discovered and from the very different technology needed to make and detect it. NASA's own wording for microwaves is telling: they are the higher frequency end of the radio spectrum, distinguished mainly by the technologies used to access them. Different textbooks put the edges in slightly different places, and the X-ray and gamma-ray bands overlap in energy: per the US EPA, X-rays and gamma rays have the same basic properties, but gamma rays come from inside the nucleus and X-rays from processes outside it.

With that caveat, here is the spectrum in one table. Frequency and wavelength ranges are approximate and computed with c = f × lambda; where a NASA, FCC or ITU source gives a specific figure, it is used.

Two practical notes on the table. The FCC treats the radio frequency portion of the spectrum as the regulated resource and notes that only 8.3 kHz to 275 GHz is currently allocated to radio services. And the X-ray row's wavelength range, 0.03 to 3 nm, is NASA's; medical and industrial X-ray tubes routinely produce photons above 41 keV, which is part of why the X-ray and gamma-ray labels overlap.

Radio waves

Radio waves have the longest wavelengths in the spectrum. NASA describes them as ranging from the length of a football to larger than the planet. They are made whenever charges oscillate: in a transmitter, an electronic oscillator drives an alternating current into an antenna, and the antenna radiates. At the receiver, the passing field drives a tiny current in another antenna, which a tuned circuit selects and amplifies.

Nature makes radio waves too. NASA notes that Karl Jansky at Bell Labs showed in 1932 that objects in space radiate radio waves, and that radio emissions from Jupiter's ionosphere have wavelengths of about 15 metres. Radio astronomy has one big advantage: sunlight, cloud and rain do not block it. It also has a resolution problem. Because the waves are long, a dish must be huge to see fine detail; NASA notes that the 64 metre Parkes dish cannot give an image any sharper than a small backyard optical telescope. The fix is to combine dishes. The Very Large Array in New Mexico uses 27 antennas spread in a Y up to 36 km across so that together they behave like one enormous telescope.

Microwaves

Microwaves are the upper end of the radio range, roughly 300 MHz to 300 GHz, where wavelengths shrink from a metre to a millimetre. At these sizes, antennas become small enough to fit in a phone and dishes become practical for focusing beams, which is why this band carries most modern wireless traffic.

NASA's microwave page lists the sub-bands remote sensing scientists use. C-band penetrates cloud, dust, smoke, snow and rain to image the surface. L-band, used by GPS receivers, can pass through forest canopy. Most communication satellites use C, X and Ku bands to reach ground stations. Microwave ovens use waves about 12 centimetres long to make water and fat molecules rotate, and the friction of that forced rotation is heat.

Radar is the classic active microwave system: it sends a pulse and times the echo. NASA's example is the radar altimeter on the Jason-2 ocean satellite, which used two frequencies, 13.6 and 5.3 GHz, to measure sea surface height to within a few centimetres. Microwaves also gave cosmology its central clue. In 1965 Arno Penzias and Robert Wilson at Bell Labs found a faint noise coming equally from every direction: the cosmic microwave background, now about 2.7 kelvin.

Infrared

Infrared sits just beyond red light. We feel it as heat. NASA notes that William Herschel discovered it in 1800 by placing thermometers in each colour of a prism's spectrum and finding the temperature kept rising just past the red end. The band is usually split into near, mid and far infrared, and Earth scientists call 8 to 15 micrometres the thermal infrared because that is where the Earth's own heat radiation is strongest.

Everyday infrared is closer than people think. A TV remote uses infrared at about 940 nanometres, invisible to the eye but often visible to a phone camera. Heat lamps emit between roughly 500 and 3000 nanometres. Thermal cameras and night vision devices image warm bodies because people and animals emit mostly infrared. In astronomy, infrared passes through gas and dust that blocks visible light, which is why the James Webb Space Telescope carries three infrared instruments.

Visible light

Visible light is the narrow slice our eyes evolved to use. NASA gives the typical range as 380 to 700 nanometres, violet to red. Isaac Newton showed in 1665 that a prism bends each colour by a slightly different angle. The colour of a hot object tells you its temperature: NASA notes that the Sun's surface, at about 5,500 °C, produces more yellow light than any other colour, while a cooler star like Betelgeuse looks reddish and a hotter one like Rigel looks blue.

Visible light is also an analytical tool. Dark absorption lines in the solar spectrum are fingerprints of the elements in the Sun's atmosphere. Laser altimeters such as the one on NASA's ICESat measure the height of ice sheets, clouds and forest canopy by timing reflected laser pulses.

Ultraviolet

Ultraviolet has shorter wavelengths than violet light. NASA notes that Johann Ritter found it in 1801 by showing that photographic paper darkened beyond the violet end of the spectrum. Earth scientists split it into UV-A, UV-B and UV-C. UV-C, the most harmful, is almost completely absorbed by the atmosphere. UV-B causes sunburn and raises the risk of DNA damage; about 95 percent of it is absorbed by ozone. Astronomers use another split (near, middle, far and extreme ultraviolet), and spacecraft such as the Solar Dynamics Observatory image the Sun in extreme ultraviolet to map gas at temperatures from about 60,000 °C to more than a million degrees.

Somewhere in the ultraviolet, photons gain enough energy to ionize atoms. That threshold is the dividing line between the two halves of the radiation safety world and gets its own section below.

X-rays

X-rays have much higher energy and much shorter wavelengths than ultraviolet. NASA gives their wavelength range as 0.03 to 3 nanometres, small enough that some are no bigger than a single atom, and notes that scientists usually describe them by energy. Wilhelm Roentgen documented them in 1895. A medical X-ray works because bone absorbs more X-rays than soft tissue, leaving a shadow on the detector.

X-rays are made when fast electrons slam into a metal target and decelerate, or when electrons drop into inner atomic shells. In space they come from gas at millions of degrees: the solar corona, the accretion discs around black holes, and supernova remnants such as Cassiopeia A, where NASA reports X-ray emitting gas at about ten million degrees Celsius. X-rays do not reflect from a mirror hit head on, so X-ray telescopes use grazing incidence mirrors that deflect photons at shallow angles, the way a stone skips on water.

Gamma rays

Gamma rays have the shortest wavelengths and the highest energies. NASA lists their sources as neutron stars and pulsars, supernova explosions and regions around black holes, and on Earth, nuclear explosions, lightning and radioactive decay. Gamma-ray bursts can release more energy in ten seconds than the Sun will emit in its whole expected lifetime.

Gamma rays cannot be focused with mirrors at all. Their wavelengths are so short that they pass through the space between atoms. Detectors instead use dense crystals: a gamma ray collides with an electron (Compton scattering), loses energy, and the charged particles it sets moving are what the sensor records. The EPA notes that stopping gamma rays can take several inches of lead or a few feet of concrete.

Worked examples with numbers

These examples use only the exact NIST constants above. Round at the end, not in the middle.

Example 1: the wavelength of 2.4 GHz and 5.8 GHz Wi-Fi

Wavelength is c divided by frequency.

2.4 GHz band edge:  λ = 299,792,458 / 2.400e9 = 0.12491 m ≈ 12.5 cm
Channel 6 (2437 MHz): λ = 299,792,458 / 2.437e9 = 0.12302 m ≈ 12.3 cm
5.8 GHz:            λ = 299,792,458 / 5.800e9 = 0.05169 m ≈ 5.17 cm
6 GHz:              λ = 299,792,458 / 6.000e9 = 0.04997 m ≈ 5.0 cm

Two practical consequences follow. First, antenna size scales with wavelength. A quarter wave antenna, a common simple design, is about 3.1 cm long at 2.437 GHz and about 1.3 cm at 5.8 GHz, which is why a dual band router can hide several antennas in a small case. Second, a microwave oven runs in the same neighbourhood. The FCC designates 2450 MHz, plus or minus 50 MHz, as an ISM frequency, and 299,792,458 divided by 2.45 × 10^9 is 12.24 cm, matching NASA's "about 12 centimetres" for oven microwaves. That shared band is the reason a running microwave oven can disturb 2.4 GHz Wi-Fi nearby.

Example 2: the energy of one visible photon

Use E = 1239.84 / lambda (in nm) for electron volts, or E = h × f for joules.

Red, 700 nm:    E = 1239.84 / 700 = 1.771 eV
Green, 550 nm:  E = 1239.84 / 550 = 2.254 eV
Violet, 380 nm: E = 1239.84 / 380 = 3.263 eV

Green in joules:
  f = 299,792,458 / 550e-9 = 5.451e14 Hz
  E = 6.62607015e-34 × 5.451e14 = 3.612e-19 J
  check: 2.254 eV × 1.602176634e-19 J/eV = 3.612e-19 J

The two routes agree, which is a good habit to keep when you write this kind of calculation in code.

Example 3: the energy of one X-ray photon

0.1 nm X-ray:            E = 1239.84 / 0.1  = 12,398 eV ≈ 12.4 keV
NASA long edge, 3 nm:    E = 1239.84 / 3    = 413 eV
NASA short edge, 0.03 nm: E = 1239.84 / 0.03 = 41,328 eV ≈ 41.3 keV

A 0.1 nm X-ray photon carries about 5,500 times the energy of a green photon (12,398 / 2.254).

Example 4: the energy of one Wi-Fi photon, and why radio acts like a wave

2.4 GHz photon:  E = 4.135667696e-15 eV/Hz × 2.4e9 Hz = 9.93e-6 eV ≈ 9.9 µeV
In joules:       E = 6.62607015e-34 × 2.4e9 = 1.590e-24 J

Ratio, 0.1 nm X-ray to 2.4 GHz photon: 12,398 / 9.93e-6 ≈ 1.25 × 10^9
Ratio, green photon to 2.4 GHz photon: 2.254 / 9.93e-6 ≈ 227,000

Now count photons. A 100 milliwatt transmitter (0.1 J per second) at 2.4 GHz emits:

photons per second = 0.1 J/s ÷ 1.590e-24 J = 6.3 × 10^22

Sixty trillion billion photons a second, each with a tiny energy. At those numbers, the graininess vanishes and the classical wave picture of Maxwell's equations describes the signal perfectly, which is why radio engineers almost never talk about photons. At the X-ray end the opposite holds: detectors count individual photons and measure each one's energy.

Example 5: where ionization starts

The energy needed to pull the electron off a hydrogen atom in its ground state is about 13.6 eV. NIST lists the Rydberg energy, the constant behind that figure, as 13.605693122990 eV. Turn it into a wavelength:

λ = 1239.84 / 13.6057 ≈ 91.1 nm
f = 299,792,458 / 91.2e-9 ≈ 3.3 × 10^15 Hz

So photons need a wavelength shorter than about 91 nm, deep in the ultraviolet, to ionize hydrogen. Other atoms and molecules ionize at somewhat lower energies, which is why the boundary between ionizing and non-ionizing radiation sits somewhere in the ultraviolet rather than at a single sharp line. Compare that with Example 4: a Wi-Fi photon has about one millionth of the energy needed. No number of them acting one at a time can ionize an atom; what they can do, at high enough power, is heat.

A calculator you can run

The same arithmetic in a short Python script, useful for checking a datasheet or a homework answer:

# spectrum.py: convert between frequency, wavelength and photon energy
C = 299_792_458            # m/s, exact (NIST CODATA)
H = 6.62607015e-34         # J/Hz, exact
E_CHARGE = 1.602176634e-19 # C, exact (J per eV)

def from_frequency(f_hz):
    wavelength_m = C / f_hz
    energy_j = H * f_hz
    return wavelength_m, energy_j, energy_j / E_CHARGE

def from_wavelength(lam_m):
    return from_frequency(C / lam_m)

for label, f in [("AM 1 MHz", 1e6), ("FM 100 MHz", 100e6),
                 ("Wi-Fi 2.437 GHz", 2.437e9), ("Wi-Fi 5.8 GHz", 5.8e9)]:
    lam, ej, ev = from_frequency(f)
    print(f"{label:16s} wavelength {lam:10.4g} m  photon {ev:.3g} eV")

for label, lam in [("TV remote 940 nm", 940e-9), ("green 550 nm", 550e-9),
                   ("X-ray 0.1 nm", 0.1e-9)]:
    f, ej, ev = (C / lam,) + from_wavelength(lam)[1:]
    print(f"{label:16s} frequency {f:10.4g} Hz photon {ev:.4g} eV")

Run it with python3 spectrum.py. You should see roughly 300 m for 1 MHz, 3 m for 100 MHz, 12.3 cm for 2.437 GHz, 1.32 eV for the remote, 2.25 eV for green and 12,400 eV for the X-ray.

Why the atmosphere blocks some bands and passes others

Absorbing gases and atmospheric windows

The Sun emits across the whole spectrum, but very little of the high energy end reaches the ground. NASA names the main absorbers as water vapour, carbon dioxide and ozone, with other gases playing smaller parts. Each molecule absorbs at particular wavelengths that match its electronic, vibrational or rotational energy levels. Where none of the common gases absorbs strongly, radiation gets through. NASA calls those regions "atmospheric windows".

Walking the spectrum from the top:

What that means for astronomy

The pattern explains where telescopes live. Optical and radio telescopes work from the ground because those windows are open. Infrared telescopes work from high, dry mountains for the partial windows, and from space for everything else; the James Webb Space Telescope's infrared instruments sit far from Earth's warm, wet air. Ultraviolet, X-ray and gamma-ray observatories must fly above the atmosphere. NASA's examples include GALEX for ultraviolet, Chandra and Hinode for X-rays, and Fermi and Swift for gamma rays. NASA's summary is plain: instruments have to be positioned above Earth's energy absorbing atmosphere to see higher energy sources.

What that means for satellites and links

Communication satellites have the opposite problem. They need a band that passes through the whole atmosphere, cloud and rain included, reliably. NASA notes that most communication satellites use C, X and Ku bands, and that L-band (the GPS band) penetrates forest canopy. Lower microwave frequencies cope better with rain; higher ones offer more bandwidth and smaller dishes but suffer more from water in the path. Satellite operators trade those effects against each other, and they also need a frequency the ITU and national regulators will let them use. Our satellite internet explainer covers how low Earth orbit constellations, terminals and gateways put this into practice, and our GPS and GNSS explainer covers the navigation signals in L-band.

How the radio spectrum is allocated and licensed

The radio part of the spectrum is a shared, finite resource. Two transmitters on the same frequency in the same place interfere. So it is managed in layers: an international treaty, national tables, and then rules or licences for each band.

Layer 1: the ITU Radio Regulations

The International Telecommunication Union's Radio Regulations are the international treaty that allocates frequency bands to radio services. According to the ITU, the current Edition of 2024 contains the texts adopted by the World Radiocommunication Conference of 1995 (WRC-95), as revised by the conferences that followed: WRC-97 (Geneva), WRC-2000 (Istanbul), WRC-03 (Geneva), WRC-07 (Geneva), WRC-12 (Geneva), WRC-15 (Geneva), WRC-19 (Sharm el-Sheikh) and WRC-23 (Dubai). The ITU says the PDF versions are free for personal use.

The Regulations split the world into three regions, and an allocation can differ between them. They also distinguish three words that are often confused:

The ITU also names frequencies by decade. ITU-R Recommendation V.431 sets the nomenclature: "band number N" runs from 0.3 × 10^N to 3 × 10^N Hz, so band 9 is 0.3 × 10^9 to 3 × 10^9 Hz, which is 300 to 3000 MHz. The radio bands most people meet are below. The metric names describe the wavelength (decimetric waves are tenths of a metre long, and so on), and the typical uses column is ours, not the ITU's.

The same Recommendation keeps counting past radio: bands 13, 14 and 15 cover 3 THz to 3000 THz, which is the infrared and the start of the visible range. It is a neat reminder that the ITU's own numbering treats the spectrum as one continuous scale.

Layer 2: national tables

Each country adopts its own table within the ITU framework. In the United States, the FCC explains that responsibility is split. The FCC, an independent agency, manages spectrum for non-federal use: state and local government, commercial, private business and personal use. NTIA, part of the Department of Commerce, manages federal use. Only bands between 8.3 kHz and 275 GHz have been allocated to services so far. The FCC's Table of Frequency Allocations combines the International Table and the United States Table and is codified at 47 CFR 2.106. NTIA publishes the well known colour chart of US allocations; its page says the chart was updated in September 2025 using data as of March 2025, and it is available as a PDF.

Layer 3: licensed, licence-exempt and ISM

Within a band allocated to a service, a regulator can authorise use in a few ways.

Licensed spectrum gives a named operator exclusive or protected rights in an area, often through an auction. Cellular networks and broadcast stations work this way. Exclusivity is what lets an operator plan a network and promise coverage.

Licence-exempt (unlicensed) spectrum lets anyone use certified equipment that follows technical rules, with no individual licence and no protection from interference. In the US these rules live in 47 CFR Part 15. Section 15.247, for example, covers operation in 902 to 928 MHz, 2400 to 2483.5 MHz and 5725 to 5850 MHz, sets rules for frequency hopping and digital modulation (a minimum 6 dB bandwidth of 500 kHz for digital systems), and caps conducted output at 1 watt for qualifying systems. Wi-Fi, Bluetooth, cordless phones and many IoT radios live under rules like these.

ISM bands are designated for industrial, scientific and medical equipment that uses radio energy for something other than communication, such as heating. The FCC's Part 18 table lists the US ISM frequencies:

Communication devices that share ISM bands must accept interference from ISM equipment. That is the regulatory version of the microwave oven problem in Example 1, and it is one reason Wi-Fi gained so much from new spectrum away from the ISM bands.

The 6 GHz expansion

On 23 April 2020 the FCC adopted rules making 1,200 megahertz of spectrum in the 6 GHz band available for unlicensed use, while protecting incumbent licensed services already there. That block runs from 5.925 to 7.125 GHz, wavelengths of about 5.1 to 4.2 cm, and it is the extra room Wi-Fi 6E and Wi-Fi 7 use for wide channels. Our Wi-Fi generations explainer covers what Wi-Fi 7 does with that room, including 320 MHz channels and multi link operation, and our Fourier transform explainer shows how OFDM splits each channel into subcarriers.

Choosing a licensing model

Where everyday technologies sit

The table below places common technologies on the spectrum. Frequencies come from the FCC, NASA and ITU sources cited; wavelengths are computed with c = f × lambda. Exact band plans vary by country, so treat the cellular rows as typical ranges rather than one country's plan.

A few rows deserve a comment. The AM and FM rows are US definitions: 47 CFR 73.14 defines the AM broadcast band as 535 to 1705 kHz, and 47 CFR 73.201 defines the FM band as 88 to 108 MHz split into 100 channels of 200 kHz. Other countries differ at the edges. The 60 GHz row comes from 47 CFR 15.255, which governs licence-exempt operation in 57 to 71 GHz. The 511 keV PET photon is the electron's rest energy, which NIST lists as 0.51099895 MeV; when an electron and a positron annihilate, each of the two photons carries that much away. Divide 1239.84 by 511,000 and you get a wavelength of about 0.0024 nm. For everything else in the US, the FCC table at 47 CFR 2.106 is the authority.

Notice how the everyday microwave rows cluster between about 0.5 and 7 GHz. That range is a sweet spot: wavelengths are short enough for small antennas, long enough to get through walls and foliage reasonably well, and the atmosphere is almost transparent there. It is why so many services compete for it, and why each new block opened for unlicensed use, such as 6 GHz, is a big event.

Ionizing and non-ionizing radiation

The physical difference

The US EPA draws the line clearly. Non-ionizing radiation has enough energy to move atoms in a molecule around or make them vibrate, but not enough to remove electrons. Its examples are radio waves, visible light and microwaves. Ionizing radiation has so much energy it can knock electrons out of atoms. It can damage tissue and DNA, and it comes from X-ray machines, cosmic particles and radioactive elements. NASA places gamma rays, X-rays and some ultraviolet on the ionizing side, and Example 5 shows why the boundary falls in the ultraviolet: that is where single photons reach the 10 eV or so needed to free an electron.

The key point is that ionization is a per photon effect. A bright radio source sends more photons, not more energetic ones. Turning up the power of a Wi-Fi router cannot make any one of its photons ionize an atom. It can, at high enough levels, deposit enough total energy to heat tissue. That is why the two families of safety rules look so different.

What RF exposure limits actually cover

In the US, the FCC sets limits for radio frequency exposure. The EPA says plainly that it does not regulate non-ionizing radiation from wireless technology, and points to the FCC. The FCC adopted its current guidelines in 1996, based on recommendations from ANSI, IEEE and NCRP, and in November 2019 it voted to keep the existing limits in place while updating some procedures.

The rules are in 47 CFR 1.1310 and use two measures:

Specific absorption rate (SAR) is the rate at which body tissue absorbs RF energy, in watts per kilogram. It applies from 100 kHz to 6 GHz. For the general public, the whole body average limit is 0.08 W/kg, and the peak spatial average is 1.6 W/kg over any 1 gram of tissue. For occupational exposure, where workers are trained and can control their exposure, the limits are 0.4 W/kg whole body and 8 W/kg peak. The FCC's RF safety page notes that 1.6 W/kg is the public limit for mobile phones.

Maximum permissible exposure (MPE) limits are given as power density and field strength for antennas and base stations, and are the only method used above 6 GHz. They are averaged over time: exposure can briefly exceed the limit as long as the average over the specified averaging time does not. The table in 47 CFR 1.1310 sets, for example:

Worked through: at 900 MHz the general population limit is 900/1500 = 0.6 mW/cm². At 2.4 GHz and 5.8 GHz it is 1.0 mW/cm², which is 10 W/m².

What these limits do not cover is just as important. They are thermal limits for RF fields, set with a safety margin. They are not limits on ionizing radiation; they are not limits on ultraviolet or laser exposure, which have their own standards; and they say nothing about data, privacy or security. On health research, the EPA's summary is that studies have not shown an association between cell phone use and cancer that would allow a causal conclusion, while research on long term low level exposure continues. Near powerful transmitters such as radar or tower mounted equipment, RF energy can cause burns, which is why access warnings around towers matter.

Outside the US, limits are set nationally. Many regulators point to international guidance from the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and to the World Health Organization's work on electromagnetic fields, but the numbers and the way compliance is tested differ, so check your own national regulator for the limits that apply to you.

Ionizing radiation in one paragraph

On the ionizing side, the EPA describes the main forms. Alpha particles are heavy and stopped by skin, but harmful if inhaled or swallowed. Beta particles are lighter and stopped by clothing or thin aluminium. Gamma rays and X-rays are photons; gamma rays are the most penetrating and can need several inches of lead or a few feet of concrete. X-rays come from outside the nucleus and are generally lower in energy than gamma rays. Medical uses rely on careful dose management, and the same physics that makes ionizing radiation hazardous makes it useful in cancer treatment, which NASA also notes.

Step by step: see which part of the spectrum your devices use

You can check where your own Wi-Fi sits, what your regulatory domain allows, and how those frequencies translate to wavelengths. None of these commands transmit anything new or change settings unless noted.

1. Find your current Wi-Fi frequency on Linux

# List wireless interfaces
iw dev

# Show the current link, including the frequency in MHz
iw dev wlan0 link

Replace wlan0 with the interface name from iw dev. A line such as freq: 5180 means you are on channel 36 in the 5 GHz band. Convert it:

python3 -c "print(299792458/5180e6*100, 'cm')"
# about 5.79 cm

2. See every frequency your card supports

iw list | grep -E 'MHz \['

Each line shows a frequency, its channel number, the maximum transmit power and flags that show whether the channel is disabled or restricted where you are. Which channels are enabled depends on your regulatory domain.

3. Check the regulatory domain

iw reg get

The output lists frequency ranges with maximum bandwidth and power for the country your system believes it is in. This is the national table, turned into data your kernel enforces. The Linux wireless documentation describes iw reg set followed by an ISO 3166 two letter country code, or a country= line in the wpa_supplicant configuration, as the way to set it. Set the country you are actually in. Do not pick a different one to unlock channels; transmitting outside your country's rules can interfere with radar and other licensed services and is unlawful.

4. On macOS

system_profiler SPAirPortDataType | grep -E 'Channel|PHY Mode'

The channel line shows the band (2 GHz, 5 GHz or 6 GHz) and the width.

5. Scan neighbouring networks without connecting

nmcli -f IN-USE,SSID,CHAN,FREQ,SIGNAL dev wifi list

This lists nearby networks with channel and frequency, which helps you see congestion in the 2.4 GHz band and pick a cleaner 5 or 6 GHz channel on your own router.

6. Turn the numbers into a band map

Feed the frequencies you found into the spectrum.py script above. You now have, for your own environment, the same three numbers this post is about: frequency, wavelength and photon energy.

The security angle, briefly and honestly

The spectrum matters to security in a few concrete ways, and it is easy to exaggerate them, so here is the short version.

Radio links are shared media. Anything sent over Wi-Fi, cellular, Bluetooth or a satellite link can be received by anyone in range with suitable equipment. The protection comes from encryption and authentication in the protocol, not from the radio. That is why WPA3, cellular authentication and encrypted satellite links matter, and why you treat any network you do not control as hostile.

Radios run software. A Wi-Fi card, a cellular modem, a satellite terminal and a GPS receiver each run firmware and drivers, and the access points and phones around them run large software stacks built partly from open source components. Vulnerabilities in that software are far more common than anything exotic in the physics, and the fix is the same as anywhere else: know what is inside, update it, and rebuild what you ship.

Electronics leak energy. Computers and cables emit unintended electromagnetic signals as they switch. NIST's glossary defines TEMPEST as the investigation, study and control of unintentional compromising emanations from telecommunications and information systems equipment. In practice, defending against this means shielding, filtering, separation distances and equipment standards in high assurance settings; for most organisations, it is a niche concern next to patching and access control. We describe it here only so the term makes sense when you meet it.

Interference is a physical attack surface. Jamming and spoofing of radio signals, including satellite navigation, are documented threats. Our GPS and GNSS explainer covers how navigation signals work and how receivers cope.

Common mistakes

Treating band names as physical boundaries. "Microwave" and "radio" are the same physics; so are hard X-rays and gamma rays of the same energy. The labels describe technology and origin, not a change in nature.

Mixing units mid calculation. GHz with metres and nanometres with joules produce answers off by factors of a billion. Convert to SI first (Hz, m, J) or use the 1239.84 eV·nm shortcut consistently.

Confusing power with photon energy. A strong radio transmitter sends more photons, not more energetic ones. Ionization depends on photon energy; heating depends on power absorbed.

Assuming "radiation" always means ionizing. Radio, microwaves and visible light are radiation too. The word on its own says nothing about risk.

Assuming higher frequency always means better wireless. Higher frequencies offer more bandwidth but shorter range and poorer penetration through walls, rain and foliage. Each band is a trade off.

Thinking unlicensed means unregulated. Part 15 and its equivalents set power, bandwidth and behaviour rules, and devices must be certified. Unlicensed only means no individual licence and no protection from interference.

Changing the regulatory domain to unlock channels. It can cause interference with radar and licensed services and is unlawful in many countries.

Quoting one country's band plan as universal. Allocations differ by ITU region and by country. Always name the jurisdiction.

Using a phone's SAR value as a ranking of safety. SAR is measured under standard test conditions at maximum power to show compliance with a limit. It is not a measure of typical exposure in use.

FAQ

What is the electromagnetic spectrum in simple terms?

It is the full range of light, in the broadest sense. Radio waves, microwaves, infrared, visible light, ultraviolet, X-rays and gamma rays are all electromagnetic waves travelling at the same speed. They differ only in frequency, which sets their wavelength and the energy of each photon.

What is the order of the electromagnetic spectrum from lowest to highest frequency?

Radio waves, microwaves, infrared, visible light (red to violet), ultraviolet, X-rays, gamma rays. Wavelength runs the other way, from longest (radio) to shortest (gamma), and photon energy rises with frequency.

How do you calculate wavelength from frequency?

Divide the speed of light, 299,792,458 m/s, by the frequency in hertz. For 2.4 GHz Wi-Fi, 299,792,458 divided by 2.4 × 10^9 is 0.125 m, or 12.5 cm.

How do you calculate the energy of a photon?

Multiply the Planck constant, 6.62607015 × 10^-34 J/Hz, by the frequency. In electron volts, divide 1239.84 by the wavelength in nanometres. A 550 nm green photon has about 2.25 eV.

Is Wi-Fi radiation ionizing?

No. A 2.4 GHz photon carries about 10 micro electron volts, roughly a millionth of the energy needed to ionize an atom. Wi-Fi is non-ionizing radiation; at high enough power, radio energy can heat tissue, which is what exposure limits are designed to prevent.

Why can microwaves heat food but Wi-Fi does not cook you, if they use similar frequencies?

Power and distance. A microwave oven concentrates hundreds of watts inside a metal box that reflects the energy back into the food. A Wi-Fi radio transmits a fraction of a watt in all directions, and the energy spreads out and weakens with distance.

Why do X-ray and gamma-ray telescopes have to be in space?

The atmosphere absorbs X-rays and gamma rays before they reach the ground. NASA's X-ray and gamma-ray observatories, such as Chandra, Fermi and Swift, orbit above the atmosphere for that reason.

Who decides which frequencies Wi-Fi and phones can use?

Internationally, the ITU Radio Regulations allocate bands to services. Each country then sets its own table and rules. In the US, the FCC handles non-federal spectrum, including Wi-Fi rules in Part 15 and cellular licences, and NTIA handles federal spectrum.

What is an ISM band?

A band designated for industrial, scientific and medical equipment that uses radio energy for non communication purposes, such as microwave ovens. In the US, 2450 MHz plus or minus 50 MHz and 5800 MHz plus or minus 75 MHz are ISM bands, and Wi-Fi and Bluetooth share them under licence-exempt rules.

What is the difference between X-rays and gamma rays?

Both are high energy photons. The EPA explains that gamma rays originate inside the atomic nucleus, while X-rays come from processes outside it, and that X-rays are generally lower in energy. Their energy ranges overlap, so the name tells you the source more than the energy.

Where Safeguard fits

Safeguard does not do physics, measure radiation, analyse spectra, manage radios or assess RF exposure, and nothing in this post depends on it. Safeguard's models are single purpose security models. The connection is indirect: the devices that use the spectrum, from Wi-Fi access points to satellite terminals and cellular modems, run firmware and software stacks built in part from open source components, and the tools teams write around them (network management apps, provisioning scripts, telemetry pipelines, signal processing services) pull dependencies like any other project. Safeguard helps with that software: it scans repositories, container images and SBOMs for vulnerable components, shows which projects contain a given component and version, and can open fix pull requests.

The physics of the spectrum does not change, but the software that drives every radio does, and a rebuilt network tool or gateway image pulls its open source dependencies again each time. The section below describes how Safeguard closes the gap for the software you build and ship; its fork and patch flow is early access.

Fix it without waiting for upstream

Whatever the vulnerability, most of the risk sits in two gaps. The first is the time between a fix being published and that fix actually running in your production builds. The second, and worse, is the time before any fix exists at all, while a maintainer works out a patch and a release schedule. Waiting through both gaps is a habit, not a requirement. Safeguard gives teams three ways to close them.

1. Autonomous remediation: pull requests, not tickets

When a fixed version exists, Safeguard's Griffin models open a tested, policy gated pull request that moves the affected dependency to the safe version. You can run it on one repository or in bulk across every project that carries the vulnerable version. Remediation covers eight categories, including dependencies, licenses, security posture and findings. Reachability analysis ranks the vulnerable code paths your application actually calls first, so the first pull requests you merge are the ones that matter most. Autonomous remediation ships switched off, with every category set to Manual, so your team decides how much runs on its own and what still needs a human approval.

2. Gold Open Source: the safe version in seconds, free

gold.safeguard.sh is free and needs no login. Search any CVE or package to see affected and fixed versions, CVSS, EPSS, CISA KEV status and a recommended safe version, or use the "Am I affected?" checker. The same data is available through a free read API, README status badges and RSS threat feeds. The safeguard-sh/gold-scan-action GitHub Action can fail a build on a known exploited vulnerability, on any known vulnerability, or never, so a version you have just removed cannot quietly come back in next week's merge.

3. Gold Registry: start from components that are already fixed

registry.safeguard.sh provides hardened, drop in replacements for upstream packages and container images. Every Gold artifact is held continuously to zero exploitable critical or high CVEs, zero malware and a completed license check, with SLSA L3 or higher provenance, a Sigstore signature and a public evidence page you can hand to an auditor. An artifact loses its Gold badge if it regresses. Teams can point their own package directory at the registry so hardened equivalents resolve without changing how developers install anything.

When upstream has no fix yet, or will not ship one, Safeguard first works to get the fix accepted upstream. If the issue is severe or upstream will not move, Safeguard builds a fork, fixes the vulnerability there and publishes a hardened Gold build, so you are not stuck on a maintainer's release schedule. You can also request a hardened, compatibility tested build of a specific package. This fork and patch flow is in early access today, so ask which packages it covers before you plan around it.

One honest limit: when the vulnerable code lives in a vendor appliance, a hosted service or an operating system runtime you do not build yourself, the fix still has to come from that vendor or distribution. The steps above apply to the open source dependencies and container images you build and ship, which is where most teams carry the largest and least tracked share of their exposure.

See how Safeguard would help

The Safeguard CLI demo is a real terminal session: you ask it to fix a vulnerable dependency, it reads the finding and the advisory, asks before each command, edits the manifest, runs the install and the tests with your approval, and pushes nothing. The same flow applies to a network management tool or a signal processing service you maintain.

Click through it at your own pace, or open it full screen.

The local runner demo sets up a runner on your own machine, approves it with a device code, then scans from the terminal. If you build gateway images or radio tooling on a build host you control, this is how scans run there.

Click through it at your own pace, or open it full screen.

The Claude demo connects Claude to the Safeguard MCP server with OAuth and chooses which tools Claude may use. Engineers who already ask an assistant about channel plans or driver options can ask, in the same place, which of their projects carry a vulnerable component.

Click through it at your own pace, or open it full screen.

Quick check

Sources

• NASA Science: Introduction to the Electromagnetic Spectrum
• NASA Science: Anatomy of an Electromagnetic Wave
• NASA Science: Radio Waves
• NASA Science: Microwaves
• NASA Science: Infrared Waves
• NASA Science: Visible Light
• NASA Science: Ultraviolet Waves
• NASA Science: X-rays
• NASA Science: Gamma Rays
• NIST CODATA: speed of light in vacuum
• NIST CODATA: Planck constant
• NIST CODATA: Planck constant in eV/Hz
• NIST CODATA: elementary charge
• ITU: Radio Regulations
• FCC: Radio Spectrum Allocation
• NTIA: United States Frequency Allocation Chart
• eCFR: 47 CFR 18.301, ISM operating frequencies
• eCFR: 47 CFR 15.247, operation within 902 to 928 MHz, 2400 to 2483.5 MHz and 5725 to 5850 MHz
• eCFR: 47 CFR 1.1310, radiofrequency radiation exposure limits
• FCC: FCC Opens 6 GHz Band to Wi-Fi and Other Unlicensed Uses, 23 April 2020
• FCC: Radio Frequency Safety
• US EPA: Radiation Basics
• US EPA RadTown: Non-Ionizing Radiation from Wireless Technology
• NIST CSRC Glossary: TEMPEST
• NIST CODATA: Rydberg constant times hc in eV
• NIST CODATA: electron mass energy equivalent in MeV
• ITU-R Recommendation V.431-8: Nomenclature of the frequency and wavelength bands used in telecommunications
• eCFR: 47 CFR 73.14, AM broadcast definitions
• eCFR: 47 CFR 73.201, numerical designation of FM broadcast channels
• eCFR: 47 CFR 15.255, operation within the band 57 to 71 GHz
• NASA: Global Positioning System, Space Communications and Navigation
• Linux wireless documentation: iw

Last reviewed: 23 October 2026. Band plans, exposure rules and allocation charts change; check the regulator for your own country before relying on a specific frequency or limit.