02Radio in plain words

The short version: Every wireless signal around you - Wi-Fi, 5G, a taxi dispatcher, a plane's transponder - is a radio wave, and the main thing that sets one apart from another is its frequency, counted in hertz. Regulators slice the spectrum into named bands from VLF to EHF, and they match each service to a band based on how waves at that frequency travel. Lower frequencies get through walls and cover more ground with fewer transmitters; higher frequencies offer far wider channels, so they carry more data, but a pane of energy-saving glass can stop them. A shipwreck hurried along the rules on who may use which slice: the US first required radio licenses in 1912, after the Titanic, and created the FCC in 1934.

What is a radio frequency, and what does "hertz" mean?

What you see: A number like "2.4 GHz" or "6 GHz" on a Wi-Fi router box, or "5G" on your phone's status bar.

A radio wave is an electromagnetic wave, the same family as light but far lower in frequency. The German physicist Heinrich Hertz is credited with having "solidified and proved the theory of electromagnetic waves," the work that inspired Guglielmo Marconi to build a wireless business.³

Frequency is how many times the wave repeats each second. One cycle per second is one hertz (Hz). Radio frequencies are large numbers, so they come with prefixes: a kilohertz (kHz) is a thousand cycles per second, a megahertz (MHz) a million, a gigahertz (GHz) a billion. A "2.4 GHz" Wi-Fi radio is putting out a wave that repeats about 2.4 billion times a second.

The unit of frequency, the hertz, is named after Heinrich Hertz.⁴ Working in a lecture hall of the technical university in Karlsruhe, Germany, he observed the propagation of an electromagnetic wave for the first time on November 11, 1886, and his measurements showed that the effect traveled at the speed of light.⁴

Wavelength is the physical length of one cycle of the wave, from one crest to the next. Frequency and wavelength move in opposite directions: the higher the frequency, the shorter the wave. That relationship is why engineers talk about "millimeter waves" for some 5G bands and why antennas for low frequencies are long while the antenna inside a phone is tiny.

The rule that ties them together: frequency times wavelength equals the speed of light, about 299,792 km per second.⁵ Dividing the speed of light by the frequency gives the wavelength, so a 100 MHz wave, in the middle of the FM band, is about 3 m (10 ft) long, and a 3 GHz wave is about 10 cm (4 in) long.

When a regulator talks about a band, it means a range of frequencies with a lower and upper edge, such as 3.7-3.98 GHz. The width of that range, 280 MHz in that example, is the band's bandwidth.²

What do VHF and UHF stand for?

What you see: "VHF/UHF" printed on a TV antenna box, a marine radio or a handheld scanner.

They are band names. US federal rules divide the radio spectrum into numbered bands, each ten times wider than the one before, and give each a symbol and an older "metric" name that hints at the size of its waves.⁶ VHF is "very high frequency" and UHF "ultra high frequency."

Source for every row: 47 CFR 2.101.⁶

Each band includes its upper limit but not its lower limit, so exactly 300 MHz belongs to VHF, not UHF.⁶ The metric names line up with the wavelengths: "metric" waves (VHF) are measured in meters, "centimetric" waves (SHF) in centimeters, "millimetric" waves (EHF) in millimeters.

Words you hear in advertising, such as "microwave" or "mmWave," are not in this table. They are industry shorthand, and they do not always match the official band edges.

Why do low frequencies reach farther?

What you see: An AM station that comes in from far away at night, or a phone that keeps one bar deep inside a parking garage when the 5G icon has disappeared.

Australia's spectrum regulator, ACMA, notes that "the way a radio wave spreads or propagates depends on its frequency," and matches services to bands on that basis.⁷ Regulators everywhere work the same way, because frequency changes how a wave behaves on its way to you.

The broad pattern: lower-frequency waves lose less energy going through walls, bend around obstacles more readily and cover a wide area from one transmitter. That is why a low-band cell site can blanket a large area, and why services that must reach everyone, everywhere, tend to sit lower in the spectrum. The building figures in the next item show the wall part of this in numbers.

Britannica describes how this works at the low end of the spectrum. Below 3 MHz, which takes in the LF and MF bands, surface waves travel along the ground "as if in a waveguide" and can cover very large distances.⁸ In the HF band, the main mode is reflection off the ionosphere: by day its low D layer absorbs HF waves, but at night the D and E layers often disappear, so waves reflect off the higher F layer and reach much farther.⁸ The broadcast section covers what that means for AM radio.

Why do high frequencies carry more data?

What you see: A phone showing "5G UW" or "5G+" and very fast downloads outdoors, which slow sharply once you step inside.

The answer is mostly arithmetic. The whole MF band, 300-3,000 kHz, is 2.7 MHz wide.⁶ The single block of C-band spectrum the FCC auctioned for 5G in 2021, 3.7-3.98 GHz, is 280 MHz wide by itself - about a hundred times the entire MF band.² The SHF band as a whole, 3-30 GHz, spans 27,000 MHz.⁶

A wider channel can carry more bits per second, the same way a wider highway carries more cars. Higher up the spectrum, there is room for wide channels; lower down, there is not. The price of that room is reach: the higher you go, the more easily walls, windows, leaves and human bodies absorb or reflect the signal. Networks handle this by stacking several bands at once and handing your phone between them, as the cellular section explains.

Why does my phone lose signal inside buildings?

What you see: Full bars on the sidewalk, one bar in the lobby, nothing in the elevator.

Every wall, floor and window between you and the cell site costs signal, and the cost climbs with frequency. Engineers measure it in decibels (dB); every 10 dB is a factor of ten in power. One industry reference summary gives these typical building penetration losses:⁹

At 28 GHz, the same summary calls concrete "effectively opaque."⁹ Treat these as ballpark figures: the page has internal inconsistencies and itself tells engineers to check the ITU's model (ITU-R P.2109) and the 3GPP 5G channel model (TR 38.901).⁹

The material matters as much as the frequency. Low-E glass - energy-saving window glass with a thin metallic coating that keeps heat in - partly reflects radio waves, and one industry summary likens the metallized coating to a Faraday cage.⁹ That summary puts its loss at about 25-35 dB at 3.5 GHz, against 5-8 dB for ordinary glass at 2 GHz.⁹ A glass office tower built to save energy can be harder to get a signal into than an old brick building with plain windows. Indoor antenna systems and small cells mounted close to the street are two ways around the problem; for the gear on poles, see https://streetcode.cold.fyi/utility-poles/.

What is "building entry loss," and how do engineers model it?

What you see: A coverage map that shows "good" service at an address where your phone struggles indoors.

Building entry loss is the signal lost between the outside of a building and a point inside it. The ITU's model for it, Recommendation ITU-R P.2109, does not give one number: it gives a frequency-dependent median that rises with the logarithm of the frequency, with coefficients that depend on building type, plus a random term to reflect how much real buildings vary.⁹ It sorts buildings into "traditional" and "thermally efficient" types. The 3GPP model used in 5G simulations makes a similar split into "low-loss" buildings (older, large windows) and "high-loss" buildings (modern, energy-efficient).⁹

Real buildings do not always fit those two boxes. A January 2024 study by Christopher R. Anderson at NTIA's Institute for Telecommunication Sciences, the US government's radio research lab, measured building entry loss at 1.7 GHz in one historic and one modern building. It found that material type and density matter a great deal, and that the P.2109 "traditional" versus "thermally efficient" categories may not capture enough construction detail.¹⁰

Outside, the city adds its own complications. In a downtown street, a signal rarely arrives on a single straight path: it reflects off buildings and bends over rooftops, and several copies of it arrive at slightly different times. This is called multipath. Phones and GPS receivers have to sort those copies out; the GPS section covers what that does to your blue dot.

GPS shows the cost in numbers. The US government's GPS website says GPS-enabled smartphones are typically accurate to within a 4.9 m (16 ft) radius under open sky, but that accuracy worsens near buildings, bridges and trees, and it lists signals reflected off buildings or walls - multipath - among the common causes of poor positioning.¹¹

Who discovered radio waves, and did Marconi invent radio?

What you see: A plaque, a museum display or a trivia question crediting one person with "inventing radio."

The honest answer is that it took several people. JSTOR Daily credits Hertz with proving that electromagnetic waves exist, and Marconi (1874-1937) with turning them into a working communications system.³ Marconi moved from Italy to England in 1896 to find funding, and in 1899 he sent signals across the English Channel.³

Other countries tell the origin story with different names at the center, and those claims are argued over by historians to this day. This guide credits each person with what the sources say they did, rather than naming one "inventor."

In Russia, the name is Aleksandr Popov. IEEE Spectrum reports that he demonstrated his radio-wave detector in a presentation at St. Petersburg University on May 7, 1895 (he later adapted it as a lightning detector), a date later declared Radio Day and still observed in Russia; Popov never sought a patent, while Marconi filed for a British patent on June 2, 1896.¹²

In the US, supporters of Nikola Tesla point to the courts: a history page at Spain's Universitat Pompeu Fabra says that in 1943, a few months after Tesla's death, the US Supreme Court overturned Marconi's patent in Tesla's favor.¹³ The ruling itself was narrower. Decided on June 21, 1943, it struck down the broad claims of Marconi's US tuning patent because earlier work by Oliver Lodge, John Stone Stone and Tesla had anticipated it; it said nothing about Marconi's original patent and did not decide who invented radio.¹⁴

What happened at Signal Hill in 1901?

What you see: References to Signal Hill in St. John's, Newfoundland, or to Poldhu in Cornwall, as the birthplace of transatlantic radio.

On December 12, 1901, Marconi and his assistant George Kemp, at Signal Hill near St. John's, Newfoundland, listened for the Morse letter "S" (three dots) being sent from a transmitter at Poldhu, Cornwall, in England.¹⁵ They used a wire antenna about 500 ft (150 m) long held up by a kite, and an untuned receiver. Marconi's diary records "Sigs. at 12:30, 1:10 and 2:20."¹⁵ The IEEE recognizes the event as an engineering milestone, though its own page notes that some historians suggest he may have heard a high-frequency harmonic of the transmitter rather than its intended signal.¹⁵

Did the Titanic lead to radio laws?

What you see: The claim, in documentaries and museum captions, that the Titanic disaster created radio regulation.

It sped it up. The Radio Act of 1912 was passed after the Titanic sank in April 1912. It required ocean-going ships to keep a radio room staffed 24 hours a day, and it put the US Department of Commerce in charge of licensing radio stations, mostly maritime and amateur ones at the time.¹⁶ It was approved on August 13, 1912.¹⁷ For the first time, anyone transmitting by radio in the US needed a government license.¹⁸

Those three ideas - a license to transmit, a federal agency in charge, and rules written partly for safety at sea - are still the core of radio law over a century later.

Why did the US need new radio laws in 1927 and 1934?

What you see: Station licenses, FCC filings and broadcast rules that all refer back to "the public interest."

The 1912 Act was written before broadcasting existed, and it broke under the strain. In 1926, Zenith's Chicago station WJAZ broadcast on a frequency it had not been assigned. On April 16, 1926, federal Judge James H. Wilkerson ruled that the government lacked authority to set operating requirements such as frequencies, and in July 1926 the Acting Attorney General agreed the law was inadequate.¹⁷ With nobody able to keep stations on their channels, Congress started over.

Federal rules today still name the 1934 Act as the authority under which the FCC runs its half of the US frequency table; the other half, for federal government users, is covered in the rules section.¹⁹

In Boston

Boston, like the rest of the 50 states, sits in ITU Region 2, and the US Table of Frequency Allocations that governs every transmitter in the city is based on the Region 2 table.¹⁹ The physics in this section is the same in Boston as anywhere else.

What Boston offers is a good place to see it. The city mixes older masonry buildings with newer glass towers, and the building entry loss research above predicts that the second kind will often be harder for an outdoor cell site to reach - especially at the higher 5G frequencies. If your phone holds a signal in an older brick building and struggles in a sealed-glass office, that is consistent with the published loss figures, which are general rather than measured in Boston.

Cape Cod has its own place in the history: the National Park Service marks a Marconi wireless station site at Wellfleet, now part of Cape Cod National Seashore, as the location of America's first wireless station.²⁰ On January 18, 1903 the South Wellfleet station took part in the first public two-way wireless communication between Europe and America, carrying messages from President Theodore Roosevelt and King Edward VII in Morse code.²⁰ The Navy closed the station in 1917 for wartime security, it was scrapped in 1920, and most of the site has since eroded into the Atlantic.²⁰

Elsewhere in the world

The band names and the physics are the same everywhere; what differs is who applies them and how. Australia's ACMA, as quoted above, plans its spectrum by matching services to the frequencies whose propagation suits them.⁷ The ITU's building entry loss model is an international recommendation, used by national regulators in coverage and sharing studies, while local building codes and energy rules decide how much low-E glass is in the windows.⁹

Radio's early history was international from the start: an Italian inventor who had moved to England for funding sent a signal from Cornwall to Newfoundland.³¹⁵ National radio laws followed in each country. In the UK, Ofcom's radio spectrum functions come from the Wireless Telegraphy Act 2006;²¹ how it and other regulators divide the airwaves is the subject of the rules section.

Report an error in this section