5G

Overview

5G New Radio (NR), standardised in 3GPP Release 15 (frozen June 2018) and first commercially launched by South Korean operators in April 2019, represents the most technically ambitious generational step in cellular history. Where previous generations primarily optimised for peak data rate, 5G is architected around three distinct use-case families: enhanced Mobile Broadband (eMBB) for extreme throughput, Ultra-Reliable Low-Latency Communications (URLLC) for mission-critical applications with sub-millisecond latency, and massive Machine-Type Communications (mMTC) for dense IoT deployments at very low cost and power.

To serve these radically different requirements, 5G NR introduces flexible numerology, millimetre-wave spectrum, massive MIMO, and a fully cloud-native core network (5GC) designed for network slicing — the ability to partition a single physical network into multiple logically independent virtual networks, each optimised for different service characteristics.

5G NR: Flexible Numerology and Spectrum

NR retains OFDM as the waveform but introduces flexible numerology through a scalable sub-carrier spacing (SCS) parameter μ. Sub-carrier spacing is defined as 15 × 2μ kHz, ranging from 15 kHz (μ=0, same as LTE, for sub-1 GHz bands) to 120 kHz (μ=3, for mmWave), and even 240 kHz for reference signals. Higher sub-carrier spacing means shorter OFDM symbol duration and shorter slot duration — enabling the lower latency required for URLLC — but requires larger guard intervals. NR thus adapts its time-frequency grid to spectrum and use case simultaneously.

5G operates across two frequency ranges:

• FR1 (Sub-6 GHz, 410 MHz – 7.125 GHz): Channel bandwidths up to 100 MHz. Good propagation, coverage comparable to 4G. The workhorse of most 5G deployments. Includes the mid-band "sweet spot" at 3.5 GHz (n77/n78), used extensively in Europe and Asia.
• FR2 (mmWave, 24.25 GHz – 52.6 GHz): Channel bandwidths up to 400 MHz (up to 800 MHz with carrier aggregation). Enormous capacity but severe range and penetration limitations — outdoor line-of-sight coverage typically under 300 metres. Primarily deployed in dense urban areas, stadiums, and fixed wireless access scenarios in the US, Japan, and South Korea.

Massive MIMO and Beamforming

Massive MIMO — antenna arrays with 32, 64, or up to 256 antenna elements at the base station — is central to 5G capacity improvements, particularly in mid-band deployments. With many more antennas than active users, the base station has enough spatial degrees of freedom to simultaneously serve many users on the same time-frequency resource using multi-user MIMO, with inter-user interference controlled through sophisticated beamforming.

In FR2, electronically steered phased arrays are essential: mmWave signals experience path loss roughly 20 dB higher than 3.5 GHz signals at the same distance, and beamforming gain (focused energy toward each user) is required to overcome this. Both the base station and handset implement mmWave phased arrays that track each other through beam management procedures — a new class of radio management not required in previous generations.

5G Core (5GC) and Network Slicing

The 5G Core is a cloud-native, service-based architecture (SBA). Network functions (NFs) such as the AMF (Access and Mobility Management Function), SMF (Session Management Function), UPF (User Plane Function), and PCF (Policy Control Function) are decomposed into microservices that communicate over HTTP/2-based service-based interfaces (SBIs), following cloud-native design principles. NFs are designed to be stateless, horizontally scalable, and deployable in containers on commercial off-the-shelf hardware — a fundamental departure from the dedicated proprietary hardware appliances of 4G.

Network slicing is the signature feature of 5GC. A slice is an end-to-end logical network constructed by instantiating a subset of NFs and radio resources with particular characteristics. An operator might simultaneously run an eMBB slice for consumer broadband, a URLLC slice with strict latency guarantees for industrial automation, and an mMTC slice for IoT devices — on the same physical infrastructure, with guaranteed isolation between slices.

URLLC: Ultra-Reliable Low-Latency

URLLC targets applications such as factory automation, remote surgery, vehicle-to-everything (V2X) communications, and smart grid control — use cases where a dropped packet or delayed response has physical consequences. The technical requirements are extreme: less than 1 ms user-plane latency and packet error rates below 10⁻⁵ (one error per 100,000 transmissions).

Achieving this requires mini-slots (as short as one or two OFDM symbols, vs 14 in a standard slot), grant-free access (a device transmits without waiting for a scheduler grant, eliminating scheduling latency), aggressive HARQ retransmission, and physical-layer techniques that prioritise reliability over throughput. Full URLLC capabilities require Standalone 5G with the 5GC — they are not available in NSA deployments.

mMTC and the Internet of Things

The massive machine connectivity use case is partially addressed by two LTE-era technologies — NB-IoT and LTE-M — which are incorporated into the 5G specification and continue in 5G deployments. These provide low-power wide-area (LPWA) connectivity for billions of sensors and actuators with periodic, small-volume data transmissions and battery lives measured in years. The 5G-specific mMTC work (under the NR-RedCap — Reduced Capability — standard in Release 17) targets mid-tier IoT devices such as industrial sensors and wearables, offering NR connectivity at lower hardware complexity and cost than full 5G.

Security Enhancements

5G introduces significant security improvements over 4G. The Subscription Concealed Identifier (SUCI) encrypts the subscriber's permanent identity using the home network's public key before transmission, preventing IMSI-catchers from harvesting permanent identifiers over the air. Mutual authentication between the network and the device is now mandatory at both the 5GC and NR layers. Security contexts are established per network slice. The home network can re-authenticate the subscriber even during roaming, reducing the risk of rogue visited-network attacks. 5G also introduces security edge protection proxies (SEPPs) for signalling protection between operators.