Two very different things under one name

5G NR was designed to span an enormous frequency range, and the gap between its two halves is large enough that engineers should probably have given them separate names.

One standard, two physical realities

When 3GPP ratified 5G New Radio in Release 15 in 2018, it defined two frequency ranges inside the same specification. Frequency Range 1, abbreviated FR1, covers 450 MHz to 7.125 GHz — the sub-6 band, in common shorthand, though the ceiling was later stretched to 7.125 GHz to accommodate a few national allocations. Frequency Range 2, FR2, covers 24.25 GHz to 52.6 GHz, the territory everyone calls mmWave. Both frequency ranges carry the 5G NR label. Both run the same NR air interface. In almost every other practical respect they are different radios deployed for different purposes in different environments.

The physics of propagation is the governing fact. At sub-6 frequencies, a signal behaves much as 4G LTE does: it diffracts around buildings, penetrates walls with moderate loss, and reaches a few kilometres from a single tower. A carrier in the 3.5 GHz band — the workhorse of early 5G deployments in most markets — can light up a macrocell with geometry similar to its LTE predecessor. For an operator, the capital structure of a sub-6 deployment is recognisable: fewer sites, outdoor coverage leading indoor penetration, manageable RF front-end complexity. 3GPP's Release 15 summary describes both frequency ranges in one document, but the engineering choices they demand are largely separate conversations.

mmWave changes every assumption. Signals at 28 GHz or 39 GHz propagate in what is effectively line of sight. A human body is enough to break a link. Rain and foliage attenuate meaningfully. A single mmWave small cell covers perhaps a hundred metres in open air, and indoor propagation means room-to-room, not floor-to-floor. The compensation is bandwidth: regulators in several jurisdictions have licensed contiguous blocks of 400 MHz or more at mmWave, where sub-6 deployments typically fight over 100 MHz channels at most. The resulting throughput numbers are real — peak data rates measured in tens of gigabits per second under ideal conditions — but they are achievable only where the geometry cooperates.

What the split means for deployment

The hardware consequences of FR1 and FR2 are substantial. A sub-6 radio site is largely familiar infrastructure: a tower or rooftop, a remote radio head, an antenna array of manageable dimension. An mmWave deployment demands a dense grid of small cells, active antenna units with integrated beamforming, and careful attention to backhaul, because a small cell that cannot be backhauled to fibre is a stranded asset. The ITU's IMT-2020 requirements, ratified in 2017, set the peak and edge data rates that 5G was built to meet, but those targets implicitly assumed the deployment would match the frequency range to the use case.

Beamforming is essential at mmWave in a way it is merely useful at sub-6. Because the wavelength at 28 GHz is roughly eleven millimetres, an antenna array that fits on a small form factor can carry dozens or hundreds of elements, enabling narrow, steerable beams that compensate for the free-space path loss. Without that beam gain, mmWave links would not close at useful distances. Sub-6 deployments can use massive MIMO arrays for capacity and interference management, but the underlying link budget is not nearly as dependent on beam forming precision.

The market outcome, four-plus years into commercial 5G, reflects the physics. Sub-6 5G is the global coverage story: broadly deployed across Asia, Europe and North America, often indistinguishable in real-world use from late-model LTE. mmWave 5G remains a niche: stadium coverage, dense urban districts, fixed wireless access in specific corridors. The United States saw the most aggressive mmWave licensing, with the FCC auctioning large swaths of the 28 GHz and 39 GHz bands by 2020, and even there, outdoor mmWave coverage is sparse outside major cities. Same standard, same brand, profoundly different technology.

The resulting throughput numbers are real — peak data rates measured in tens of gigabits per second under ideal conditions — but they are achievable only where the geometry cooperates.

Read next

The rest of two lines of descent

• LineageTwo megabits, and almost nobody used itThe original 802.11 was ratified in 1997 at up to 2 Mbit/s, and it took 802.11b in 1999 to make wireless networking practical.
• Lineagea, g, n, ac, ax — a line that became unreadableThe lettered amendments carried real technical changes, and by 2018 the Wi-Fi Alliance renamed them retroactively because the letters had stopped communicating.
• LineageAdding a band, then adding widthWi-Fi 6E opened the 6 GHz band and Wi-Fi 7 widened channels and added multi-link operation, both of which are spectrum decisions before they are silicon ones.